Every global hiring pitch sounds the same until you ask the right questions. Here are three I’d ask:

• Who legally employs our people in each country?

• What will we actually pay beyond the quoted price?

• Do they have local compliance experts who can help when issues come up?

Not all EORs are created equal… Remote’s checklist covers 8 questions to ask an employer of record (EOR) before you sign.

Read The 8 Questions

The Anthropic S-1 Is Here!!!

The long-anticipated Anthropic S-1 dropped today. Looks like they are really going public soon (likely in about 3 weeks)!

The reason this IPO is so important is because it’s not just about Anthropic. Its effects will extend well beyond just Anthropic. Maybe more so than any other IPO in history:

• What do Anthropic’s numbers tell us about the broader AI trade?

• What do they mean for the chips, data centers, and power infrastructure behind the AI boom?

• What do Anthropic’s results reveal about the application layer and more traditional SaaS companies?

• Will AI take our jobs, create more jobs, or kill us all?

I read the entire Anthropic S-1. Well…I skimmed a lot of the boring stuff (there were 94 pages of risk factors!)

Below are some of the most interesting things I found 👇

Financials

Revenue

Anthropic’s revenue growth is obviously insane. Whether they IPO at $1.5T or $2T, what they have accomplished is incredible

I laughed when I saw Jamin’s revenue growth benchmark for Anthropic compared to the public software companies. So…you’re saying Anthropic has pretty good growth?

The number of customers generating at least $100K in TTM revenue has exploded. 900% NRR (see below) has definitely helped here

Net Revenue Retention (NRR)

Every CFO deeply feels Anthropic’s 900% NRR…Obviously great for Anthropic, but spending on Claude (and AI overall) has quickly become the largest expense line at every company. That’s OK if those customers are also seeing their revenue growth rate rapidly improve (or costs come down), but…not sure most of them are.

And no…900% NRR isn’t normal. And also a reminder why you should use metric medians and not averages :)

Gross Margins

Anthropic’s gross margin improvement is unprecedented…

Jamin showed two lines in his benchmark analysis - the lower one is LTM and the higher one is L6M (gross margins are rapidly improving!)

I don’t think most people thought the AI model companies could have gross margins this high.

But before you get too excited…let me explain why this isn’t necessarily apples-to-apples. Anthropic disclosed:

Model versions released in the six months ended September 30, 2026 accounted for over 80% of our revenue for the three months ended September 30, 2026.

Model development sits in R&D (not in COGS). Typical software companies are similar. Essentially all software development costs go into R&D (with the exception of sometimes a small amount of capitalized software that hits COGS). But software companies have historically built something and then could benefit for several years from that original development. Anthropic is saying that 80% of their revenue is dependent on the very latest models.

It raises the question of how much ongoing R&D is needed to sustain today’s revenue. R&D spend is very high at Anthropic…

For illustrative purposes…below is what gross margins would be if you included 50% of R&D in COGS (arbitrary amount to show sensitivity).

With 50% of R&D included in COGS, H1 2026 gross margin would be about 55.5%. Still pretty good tho. The point is that I don’t think the stated gross margins are very comparable to most other companies given the shelf-life on models. But if they do achieve recursive model training and need less R&D to build the new models….then…maybe it doesn’t matter anyway.

Compute and infrastructure expense per dollar of revenue fell 81% from Q1 2025 to Q2 2026.

Profitability

In H1, Anthropic had positive FCF of $3.4B, which is $3.9B of operating cash flow less $482M of capex.

And before the SBC crowd comes out to say “what about SBC?!?!”….SBC was “only” $2.2B. So even after subtracting SBC, FCF was positive.

Adjusted operating margin goes from very negative to positive very quickly.

Revenue/FTE: Revenue per employee is $6.6M, up from $1.4M a year earlier 🤯

Other Interesting Learnings

Competitor Revenue Concentration Risk?

Anthropic disclosed that 5-7% of their YTD revenue comes from companies using their products in violation of their policies.

Applying the 6% midpoint to Q3 estimated ARR is ~$4B of ARR coming from current/future competitors 🤯

What does this mean? This is primarily usage from other companies using Anthropic to “distill” their models so they can improve their own AI models. In other words…this is a “future/current competitor revenue concentration risk” lol

Regular Customer Concentration Risk?

Just two customers accounted for 24% of revenue in 2025. Both were below 10% in H1 2026, while customer C accounted for 11%.

While the prior top 2 customers fell below 10% of total revenue, the top 5 customers still represent 21% of total revenue in Q3 2026.

Channel Risk

So much revenue flows through the 3 primary CSPs (AWS, Azure, and Google Cloud)

2nd Tier Auditors

If Anthropic can use a 2nd-tier audit firm (Grant Thornton) through a year before a $2T IPO, then your little software company that has <0.05% of the valuation doesn’t need to use a Big 4 firm…

I have been pushing this for a long time.

Most of you will never IPO. And the ones that do will take much longer than you expect. Many VCs push Big 4 auditors. And founders (especially first-timers) feel like it’s a signal that you’re a big deal if you have Big 4 auditors. It shouldn’t be…. They cost more money, take a lot more time, and your financials are rarely more accurate because of it.

AI Pricing

Anthropic’s revenue is generated based on two pricing models:

1. pay-for-intelligence pricing (consumption)

2. subscription-based pricing

Consumption pricing rules for most AI products.

Definitions:

Net Revenue Retention

Reflects customer expansion, contraction, and churn as of a measurement date. We calculate Net Revenue Retention by dividing the Qualified Customers’ Current Period revenue by their Base Period revenue. A Qualified Customer is a customer with at least $1 of revenue during the Base Period, excluding subscribers to our Claude Pro and Max plans, whose fixed-price subscriptions do not reflect the expansion and contraction behavior this metric is intended to capture. The Base Period is the three months ending a year prior to the measurement date. The Current Period is the three months ending as of the measurement date.

Run-Rate Revenue

I’ve heard several companies recently adopting this definition (28 days * 13)

An operating metric that refers to the estimated annualized value of our revenue-generating activity as of a measurement date, calculated as the sum of (i) our consumption-based fees during the trailing 28-day period ended on the measurement date, multiplied by 13, and (ii) our aggregate monthly subscription fees as of the measurement date, multiplied by 12. For purposes of calculating Run-Rate Revenue, consumption-based fees and subscription-based fees are derived from customer usage and active subscriptions.

Compute Commitments

Anthropic has made enormous compute commitments. It might be the right decision but it’s certainly a risk. Customers can reduce spending much faster than Anthropic can reduce its infrastructure obligations.

Weird Structure

Their voting structure is weird.

1. They created a “Long-Term Benefit Trust” which is a committee that selects 4 of the 7 board members. These are “financially disinterested trustees with expertise spanning AI safety, national security, public policy, and social enterprise”.

2. A separate founder-held share carries 50.1% of voting power on matters submitted to common stockholders. This is a single share (that controls common stock majority vote) that is controlled by 7 people. So even if the founders sold 98% of their stock, they would still have majority control (filing says they will be removed if they don’t hold >2% of their IPO-date holdings).

Final Thoughts

How Anthropic’s IPO performs and every subsequent earnings release will move entire markets (software and AI companies especially).

I have no idea if Anthropic should be valued at $1.5T, $2T, or even $2.5T. The next 6-12 months are probably going to be very bumpy for the stock.

I look forward to watching…

Footnotes:

• If you are expanding globally, check out Remote’s 8 questions to ask potential EORs (today’s sponsor)

• I just opened up 2027 sponsorships - hit me up (onlycfo@onlycfo.io)

• Subscribe and share the OnlyCFO newsletter with your teams

Subscribe now

*Nothing in this article constitutes legal, tax, or investment advice

Every global hiring pitch sounds the same until you ask the right questions. Here are three I’d ask:

• Who legally employs our people in each country?

• What will we actually pay beyond the quoted price?

• Do they have local compliance experts who can help when issues come up?

Not all EORs are created equal… Remote’s checklist covers 8 questions to ask an employer of record (EOR) before you sign.

Read The 8 Questions

Capital Allocators Need to Understand AI Economics

AI-related costs are quickly becoming (if not already) the largest expense on the P&L:

• Maybe AI features are killing your gross margins

• Everyone is using expensive frontier models (across all departments)

• Your legacy software is now embedding AI that costs more

• AI engineers are expensive

Good capital allocators (the CFO’s job but also all leaders) need to understand the data, trends, and insights around AI so they can make better decisions with the fastest growing and likely the largest expense.

I asked my favorite data story-teller, Peter Walker, to share what he is seeing in his new role at OpenRouter as Head of Insights. OpenRouter is an AI gateway and marketplace that gives you access to hundreds of LLMs through one account. So they got tons of interesting data from all the AI usage that flows through their platform

The Current State of AI Tokenomics

In January 2025, OpenRouter processed about half a trillion tokens a week. By late September 2026 it was 164 trillion. Over the same stretch, the average price paid for a million tokens fell from $0.77 to $0.14.

This sort of usage growth coupled with a massive decrease in cost is pretty rare. Finance leaders have treated AI thus far as a line item like any other, albeit with significant variance. We think that approach is out of date. AI spend will become the single largest cost for most enterprises and managing it will require an understanding of marketplace dynamics.

Most companies still buy AI the way they buy software. They pick a vendor, sometimes negotiating an annual commit, and re-evaluate at set milestones. The data says inference behaves much more like a commodity input, with falling unit costs, rising volume, and a best supplier that changes every few months.

Four forces reshaping the market

1. Supply exploded

862 models have launched on OpenRouter since the start of 2024. 362 of them launched in 2026 before October, more than all of 2025. The pace of new model creation is unlikely to flag in the coming months and years.

Sure, some of these new models will come and go without making an impact. But the number of models handling more than a trillion tokens a week went from 3 to 27 over just the last 6 months. Models have become both more general (at the frontier) and more specialized (in the line behind the frontier). Just keeping up with the changes is a full-time job.

2. Open-weight models have taken major token share

Open-weight models now carry about 60% of weekly tokens on OpenRouter but only 26% of spend. A large share of everyday work (classification, extraction, coding subtasks, agent steps) has moved to models that cost a fraction of the frontier.

About two thirds of open-weight tokens run on models from Chinese labs, which raises an obvious procurement dilemma. However, where a model was built and where it runs are separate decisions. 44% of tokens on Chinese-built models are served by US-headquartered providers, so the model can be chosen on price and quality while hosting is chosen based around privacy concerns.

Almost every model is now served by several inference providers, and the number of provider endpoints serving traffic grew from about 330 at the start of this year to 1,342. An endpoint in this chart represents a combination of model x provider (so DeepSeek V4 Flash on Coreweave, for instance).

3. Agents arrived, and agents run on cache

On February 6 this year, software agents passed humans as the main source of tokens on OpenRouter. By late September, agentic workloads made up about 131 trillion tokens a week, against 18 trillion from human users.

It’s reasonable to assume that agents will use far more tokens than humans from now on.

Beyond the volume difference, there is a crucial distinction between agents and humans in token traffic. Agents resend the same long context on every step, which makes prompt caching the most important pricing feature in AI. A cached input token typically bills at a small fraction of the normal price.

So share of cache tokens is the canary in the cost coal mine. Newer models cache far more of their agentic traffic: 88% for models launched this quarter, against 55% for models launched in spring 2025. Inference providers themselves are competing fiercely on cache rate specifically. The number of providers caching more than 90% of their agentic traffic went from 2 of 35 in January to 11 of 54 in September.

4. The best model keeps (and will keep) changing

The top lab by tokens on OpenRouter has changed five times in eight quarters, and the share for the leader fell from about 40% to 21%. At the start of 2026, Anthropic took about 80% of the combined Anthropic and OpenAI spend. By September the split was close to even.

Headlines love to focus on these battles for points of share, but it’s important to keep in mind the massive underlying growth in the ecosystem broadly. A small share of a massively growing pie can still be an economically useful market (no comments on any or all upcoming IPOs!)

The price of a token

Put those forces together and you have a market where usage and spending have decoupled. Among AI-native startups that spent at least $100 on the platform in September 2025, tokens grew 18x and spend 5.1x. Among large organizations, tokens grew 5.1x and spend 2.4x.

While the gaps may be wider or smaller, users across all segments on OpenRouter saw tokens increase faster than spend in 2026.

Many others have explained the reasons behind the explosion in tokens. But what is driving the continued decrease in price?

First, cheaper models became more capable. Certain tasks don’t need Opus 5.5 and its high list prices, they can be served perfectly well by Kimi K3 or DeepSeek V4 Flash.

Second, agentic cache rates are being optimized across most providers and given most of the new tokens are agentic that has a tumbling effect to price per token. Remember, the same model can cost very different amounts depending on which provider serves it and how much of your traffic that provider caches.

For three popular open-weight models, what customers actually paid varied 43% to 89% below the list-price equivalent AND the provider with the lowest list price was often not the cheapest to use. This market moves by the minute, not the quarter.

Third, the frontier labs (OpenAI and Anthropic) have both released highly capable but much cheaper models in recent months. Luna and Haiku are new entrants in different Pareto points that satisfy many routine tasks.

Buying on landed cost

About 77% of requests on OpenRouter go to models served by more than one provider, so most traffic involves a supplier decision for each request. In prior years, our default routing made that decision mostly on input token price. With our upgraded routing algorithm, default routing prices each request on what a finance team would call “landed cost”: the effective price after cache discounts + the cost of waiting + the cost of retries.

The retry piece is simple. A request that costs $1.00 and succeeds 80% of the time costs $1.25 per successful answer. A provider can look cheaper on paper but actually be more expensive in the end based on poor quality (expressed as fails or rate-limited responses).

This has major implications on the supply side of the market. Once traffic follows landed cost, providers will compete on speed and reliability as well as price.

What matters beyond price

Of course landed cost is not the only buying criteria CFOs should consider these days.

Data. Zero-data-retention endpoints are available for models that make up 93% of tokens on OpenRouter, including 18 of the top 20. While different organizations may make slight allowances on ZDR, by and large it is a default requirement.

Uptime. No single provider is up all the time. In samples of 60 million requests in early October, 92.8% succeeded on the first provider tried and 96.9% succeeded after automatic failover. During one capacity crunch, the 15 hardest-hit providers succeeded on as few as 18% of attempts, while requests routed through OpenRouter still completed 97.3% of the time.

Takeaways for CFOs

If AI is a market, the finance leader of today needs to treat this as a volatile input to any company model they’re building. A few suggestions that fall out of this thinking:

1. Using more tokens ≠ always spending more. Hundreds of companies on OpenRouter have successfully decreased spend on tokens while using many more tokens per month. Many more have slowed spending growth while massively expanding usage.

2. Open-weight models are not monolithic. Providers vary widely while serving the same open-weight model. Locking yourself into a single provider can have upsides and downsides across speed, quality, and reliability.

3. Guardrails are your friend. Most CFOs will insist (correctly!) on zero data retention. But what does your organization require in terms of in-region routing, spend limits by function, fallback during outages, and a number of other aspects?

4. Expect more change. Looking back even 6 months feels like a different era. The rate of change is, if anything, increasing. The best CFOs will prepare to rip up best-laid plans.

The model your company will love most may not have even launched yet. Only companies with a culture of experimentation, with a finance perspective to match, will be able to stay at the frontier in terms of applied intelligence. But it is clear now that the future will involve many kinds of models and that competition will help companies control costs.

Footnotes:

• Thanks, Peter, for the incredible data! Check out his OpenRouter blog for more insights.

• If you are expanding globally, check out Remote’s 8 questions to ask potential EORs (today’s sponsor)

• Subscribe and share the OnlyCFO newsletter with your teams

Subscribe now

*Nothing in this article constitutes legal, tax, or investment advice

Forecasting AI spend for 2027 is hard. I know I’m struggling with it.…Things are changing every couple of weeks.

My friends at Tropic help their customers manage and forecast AI spend every day. Check out their on-demand webinar (Forecasting AI Spend Won’t Sit Still) to hear their CEO and Head of Procurement share how CFOs are budgeting for and managing AI spend.

How To Forecast AI Spend

Bad Procurement is VERY Costly

Speed of execution (combined with the right strategy) matters more than ever. AI has enabled everyone to move faster so if you don’t move fast, your competitors are going to eat your lunch…

The software and tools your employees use are one of the critical things to get right to gain an advantage over your competitors. Not just to save money (although I love that), but to move faster.

I asked the smartest procurement guy I know, Michael Shields, to discuss how he is managing renewals in the age of AI. Love his insights here and you should definitely forward it to everyone at your company who touches procurement (CFO, FP&A, Controller, broader exec team, etc)

Stop Using The Term “Renewal”

A few years ago I was working late in the office. My boss (CFO) walked past on his way out. “EOQ volume pretty high?” “yeah”, I replied. “We are working on over 40 renewals this quarter.”

Back then, we prided ourselves on engaging early. As a team, we met twice a month to review contracts that were expiring. We took actions to meet with stakeholders to determine our sourcing strategy.

• Did we want to kick off an RFP?

• What did utilization look like?

• How happy were our stakeholders with the provider?

Sure we worked on “new spend” requests as well but as a procurement team, the majority of our work was managing renewals. And our goal was to engage at least 90 days out….giving us plenty of time to casually look at other sources, but then in most scenarios, negotiate an optimal contract with the incumbent (even if a few went to the last minute to secure optimal pricing).

At the time, this worked great. Engaging 90 days out vs 30 days out led to meaningful results.

The problem is that this model doesn’t work in this ‘exciting’ era of AI.

Renewal Rates Are Way Too High

My anonymous CFO friend wrote an article last month called I’m Churning. It was one of those things where I read it and I suddenly felt very validated.

Do I believe we should all send emails to our supply base with cancellation notices?

No.

Did OnlyCFO showcase a bunch of compelling reasons why our processes need to be adjusted?

Absolutely.

OnlyCFO listed 10 reasons to churn. My top 3 from that list were:

1. Exponential growth in competition

2. Decline in switching costs

3. AI has changed what I value.

Perhaps unsurprisingly, software renewal rates are finally coming down. Median gross revenue retention has slid from 90% in 2022 to 84% today, and 4 of those points came in the last year alone (Source: Benchmarkit). Even the top quartile fell from 95% to 91%. This seems to be proof that buyers are waking up.

Sort of.

Folks, I’m here to tell you that 84% is still waaaaaay too high.

This isn’t meant to be a hot take. Nor am I advocating for the downfall of SaaS. Before you break out the pitchforks, hear me out.

In 2008, a much younger Michael Shields was sitting in his first ever procurement class at BYU. The professor had a slide up with Martin Christopher’s famous idea, “Companies don’t compete against companies. Supply chains compete against supply chains.”

Then it was inspiring. Now I basically hold that idea as scripture. Here’s why:

Just like good employees can be a competitive differentiator for the companies they work for, an optimized supply base (delivering strong ROI) can be a competitive advantage for companies. That’s the whole point of spending money to begin with. Companies spend money to achieve business objectives. Said another way, if you can spend money better than your competitors, you have a leg up.

However, the assumption that the supplier that you chose 3 years ago (or 2 years ago, or even 12 months ago) is still the best one is very dangerous given how much has changed in such a short period of time.

I’ll add to / reinforce OnlyCFO’s list:

• The competitive landscape has changed. Legit (and not so legit) competitors have popped up. Some of the newcomers are having quite a bit of success. Disruption is real - even for sticky solutions like ERPs.

• The product you’ve been using has likely evolved significantly (and if it hasn’t, that might be another problem altogether). Some of these changes mean broader capabilities. Instead of 2 or 3 providers, you might be able to consolidate to one.

• Needs have likely changed. Buyers might think they need the same thing but if the problem has shifted, then the needs have shifted and the decision criteria has shifted.

• The pricing has likely changed. And I’m not just implying a 5% uplift either. We’re talking big changes (especially as everyone races to move to consumption models). This isn’t always a bad thing but needs to be factored in.

• Buyer expectations have changed. I’m not saying our expectations are always based in sound reason (because it’s often not) but AI has made it so we want more…. for less.

• Build vs buy is center stage. Build doesn’t always make sense (in fact, it often doesn’t). But it’s worth considering. More on that in a moment.

I could go on but if you still need more proof, listen to this stat I heard at Gartner’s procurement conference (September 2026):

“79% of buyers regret their technology purchase within the first year.”

That’s not just a mic drop moment. That stat is damning.

Here is what it means:

We are getting purchases wrong the (vast) majority of the time.

And remember that competitive advantage I talked about earlier (optimizing the supply base)? If we are getting our purchase decisions wrong 79% of the time and our competitors are doing better than we are, they have a massive advantage.

A Few Recommendations

The first step is the easiest.

STEP 1: STOP USING THE TERM “RENEWALS”

Instead of “I’m working on 40 renewals this quarter.”

Try, “I have 40 expiring contracts this quarter.”

This isn’t just replacing terminology. It’s a mindset shift. The word ‘renewal’, by definition, implies that you will renew. And while there still is a chance you’ll renew (more on that in a moment), it should be an option, not the default. Especially in this world of crazy change.

Expiring contracts, on the other hand, means your contract is ending and you need to figure out the best course of action.

STEP 2: DON’T JUST TRACK CONTRACT END DATES

Admittedly this is a bit of a process change. It certainly was for me.

Prior to the world of AI, the bottleneck was limited data. Even if you had data, you might not be able to leverage it all. It wasn’t actionable because of constraints such as time and processing.

AI has blown all that up. Now the limiting factor is accessing the right data.

Instead of reviewing contract end dates, I have my system set up (in my case Tropic) where signals are pushed to me. The graphic below illustrates but essentially I’m hooked up to internal systems for relevant data, as well as other data relevant to my suppliers (and the associated landscape). This is what drives my actions.

These are all signals that can be actioned regardless of when the contract is expiring. That being said, while many of these signals are being pushed to me, I love pulling relevant data at strategic points. A good example would be at the midpoint of the contract. This is a great time to send off a pulse survey to collect CSAT, do a quick survey of the competitive space, and pull a quick benchmark.

STEP 3: CHALLENGE ASSUMPTIONS AND BIAS

I imagine that at this point in the article, I’ve offended a host of CS leaders and Account management leaders (the ones who typically own the renewals). Perhaps this is where I redeem myself a bit.

Remember that stat from earlier? The one where companies are regretting the vast majority of their purchasing decisions?

(Who am I kidding, of course you remember, it’s probably haunting you like it’s haunted me for the past few weeks)

Here are common mistakes I’m seeing:

1. Assuming the existing supply base is optimized

We’ve now discussed this at length. “If it ain’t broke, don’t fix it” might mean you are using your BlackBerry when others have the latest smartphones (did I just date myself?)

2. Assuming the existing supplier is outdated or the wrong supplier

The opposite mistake is just as expensive. Every week there’s a new AI-native tool promising to replace your incumbent, and it’s easy to assume newer means better. Sometimes it is. And while some switching costs have come down (like switching between tools), the switching costs associated with a bad decision are as high as ever. Think of the lost productivity when you switch and then 6 months later you realize you made a mistake (and then perhaps boomerang back). Six months of unoptimized usage… possibly at scale.

True, some companies are getting displaced. But plenty of companies - those ones we’ve been using for years - are turning the corner, investing in their product, and maintaining their status as a market leader. So the answer might be a better contract for the proven incumbent rather than a new vendor.

A related note, will some of these hot start ups last or burn out? Not gonna lie, the news about Pulley (Carta alternative) shook me. I’ve seen a few demos and even talked to their customers. They seemed like a VERY viable alternative. Everything about them seemed legit. And it was nice to have a strong competitor in the space. And then, almost overnight they were gone. Thankfully they “exited” in a way that didn’t leave customers stranded.

But it could have been worse.

3. Buying when you should build.

The build math changed. A lot of buying habits haven’t. According to McKinsey, 32% of companies have already passed on at least one software purchase because they could build it with AI coding tools. The thing you’re about to renew might be a few weeks of work for someone on your team. Just vibe code it!

4. Building when you should buy.

Woah, hold your horses, you vibe-coding vigilante.

Building is often cheap. Owning is often expensive. Retool found 60% of builders shipped software outside IT oversight in the past year (talk about a governance problem waiting to happen…) The old iceberg analogy is grossly overused but I honestly can’t think of a better analogy in this situation. So many of the costs are not visible and sometimes really hard to measure. Some of the less obvious costs are often understated or forgotten all together. Things like maintenance costs, engineering enhancements and training costs. But let’s not forget the impact and trade-off costs. Forcing users to utilize a subpar tool can dramatically affect the intended results. Sure you might save a few bucks (or maybe not) but if the results you were looking to achieve also dropped, your total cost of ownership can easily skyrocket.

Case in point:

Earlier this year, I attended a conference and felt like I was falling behind because I didn’t have nearly enough agents. So I spent the whole flight home (cross country) building one. I was so proud. But looking back, my baby was ugly. I dealt with it for 5 months before looking at the market. Someone had built a tool that did what my agent did. Except their tool was WAAY better. And when I measured the costs, it was WAAAY cheaper. I killed my agent and bought the tool.

5. Assuming your needs are the same.

Turns out business needs are often changing faster than pricing models for Claude (okay maybe that’s a bit of an exaggeration but the point still holds). You have a renewal expiring contract and the default is to leverage a tool (like Tropic!) to find out who the competitors are. Be careful, you just skipped a couple steps. First you need to go back to the business and ask the question: “What is the business problem we are trying to solve?” Then define decision criteria that will help you solve for the problem in an objective way ( otherwise you are essentially doing the same thing as going grocery shopping without a list… except with a 5,6 or 7- figure budget). Then, and only then, do you determine which suppliers are well suited to solve the problem.

6. Assuming your contract length should be long (or short).

Everyone has a default. Some teams lock in three years because that’s how they get the discount. Others are pushing everything month-to-month because “the market’s moving too fast.” You are guessing. On one hand, you are paying a premium for flexibility that you’ll never use and you’re eating uplifts and running the same negotiation over and over (ask me how I know). In other situations, you go long on a tool in a category that’s changing every quarter and you might be stuck with something obsolete in eighteen months.

Term length should not be a policy. It’s a strategic decision based on….Data.

Assess Your Data

To be successful here you need strong data. Seriously, this is the foundation. And it’s more important than ever. Do a quick analysis of your data:

I know I’ll offend a few people, but ChatGPT doesn’t have all of the information. When I’m working on my spend strategy I’m pulling in data from internal sources and external sources. Sure you can give Chat access to your internal data. Yes, you want it pulling in publicly available data. But you also need data that is not publicly available (and yes, you might need to pay for this).

Full disclosure, I run procurement at Tropic (the sponsor of this newsletter). Which means I’m customer zero. And the thing it gives me that I couldn’t get anywhere else is that third bucket. The data that isn’t public. Benchmarks on what other companies are actually paying. Negotiation playbooks built on billions in spend and thousands of contracts. Plus the real life experience of a services team that’s been in the room for these negotiations. That’s not something you can just google (but you can access it through an MCP). And it gets pushed to me alongside my internal signals, so I’m not stitching it all together myself.

Cool right? So let’s wrap things up.

Things are changing quickly and I’ll be honest, sometimes I think it’s a scary situation to be in.

But I’m doing my best to pivot and see this as a massive opportunity. The ones that do it right will not only have a strategic advantage, but they’re more likely to adapt and survive.

And when my CFO slacks me, I can say, “I’ve got 40 expiring contracts this quarter. No surprises. I’ve been watching the signals on all of them for months. And I already know which ones need real work.”

Footnotes:

• Talk to my friends at Tropic to uplevel your procurement function (move faster and spend less)

• Thanks, Michael for the masterclass on procurement in the age of AI! He even wrote a book on procurement that just came out! check it out

• Subscribe and share the OnlyCFO newsletter with your teams

Subscribe now

*Nothing in this article constitutes legal, tax, or investment advice.

What should the finance function look like in the age of AI?

I keep thinking about this question and what it means for my people, systems, and processes. Most finance teams I talk to are being slowed down because of legacy org structures.

Join this webinar next month to hear how Rillion and Outpost have done it

Register for Webinar

The topline “revenue” metric that everyone in tech talks about is ARR (or annual recurring revenue). But ARR ain’t what it used to be in a traditional SaaS model….

Many AI companies have “ARR” that is:

1. Not Annual (e.g. annualized from short periods of usage)

2. Not Recurring (e.g. one-time or unlikely to repeat)

3. Not Even Revenue (e.g. can’t recognize GAAP revenue)

Aggressive ARR definitions can create large gaps between “ARR”, GAAP revenue, and cash actually collected. A big gap between ARR and GAAP revenue doesn’t automatically mean anything is wrong. Fast growth can create a large gap. But sometimes fast growth hides the other potential issues.

When I ask teams to explain the gap between ARR and GAAP revenue, many of them act surprised that they should have to explain the difference. They have never reconciled it…

Investors are (or should be) asking more questions…

Reconcile ARR → GAAP Revenue

Performing an ARR → GAAP revenue reconciliation does two things:

1. Makes sure I don’t look like an idiot in front of the board

2. Validates that current ARR methodology is still appropriate

Few things make a CFO (or CEO with their board) lose credibility quicker than not being able to explain ARR vs GAAP revenue (and cash). And if you can regularly review this type of reconciliation then it will force you to determine if your ARR definition is drifting because of changes in the business.

Below is an example of the reconciliation I perform:

It’s obviously not going to perfectly tie out ($155K difference remaining) because we are using high-level assumptions to get reasonably close. So I would be pretty happy with only a 0.3% difference.

Timing Differences

ARR is a snapshot metric, while GAAP revenue is recognized as it is earned.

This means that companies growing really fast will report a lot more ARR than GAAP revenue. The same idea applies to companies that are very back-end loaded in their ARR bookings in the period.

In an extreme example, if a company has $1M of ARR at the beginning of the year and booked $10M ARR on the last day of the year then you will see 1,000% ARR growth. But GAAP revenue might be nearly flat because, under GAAP, they recognized at most only 1 day’s worth of revenue (so around 1/365th of ARR), assuming the contract starts on December 31st.

So a big difference between ARR and GAAP revenue isn’t necessarily a bad thing, but you should be able to explain it.

GAAP Adjustments:

There are a lot of GAAP differences from your definition of ARR. Below are a few common ones:

• Free Months: This should get spread over the contract term for GAAP, but might get ignored for ARR or just averaged into the ARR number.

• Ramps: Price-only ramps get spread across the contract for GAAP revenue while service ramps (more seats, usage, etc) get recognized as earned from additional services. ARR often does something a little different.

• SSP Allocation: If you have professional services then GAAP may require an allocation out of subscription revenue and into services (can be material especially when companies give away services for free)

• Contract Amendments: GAAP may require one-time catch-ups of revenue or prospective adjustments to the contract

• Usage-Based Revenue: For GAAP, if it is purely usage-based, then revenue is recognized as consumed but ARR might do an estimate, not include spikes in usage, take a trailing average, etc.

• On-hold Revenue: For GAAP, sometimes you have to stop recognizing revenue on a contract (usually because of significant concerns with cash collectability from a customer). But the ARR often doesn’t fall off until the end of the contract. Bad debt expense is normal, but at the point when the cash collection is doubtful you may need to stop recognizing revenue on the contract. For instance, if an annual contract billed upfront remains unpaid after six months and there is no indication the customer will pay, you should reassess whether further revenue recognition is appropriate.

• Gross vs Net: If you resell something then GAAP might have to show the net amount (the cash proceeds you keep) vs the gross selling price. But for ARR purposes you take gross. FWIW - I think GAAP and ARR should be consistent here.

• Refund/Credits: If your CSM team gives a refund for satisfaction issues then that reduces GAAP revenue, but might not impact ARR

These bridge items aren’t necessarily bad, but they can point to an aggressive non-standard ARR definition. If you reconcile and report it, then everyone will at least be on the same page about what ARR actually means.

Revenue to Cash Bridge

ARR is a hope, GAAP revenue is a judgment, and cash is a fact.

The above is particularly true for private companies. And very true for smaller private companies whose accounting practices might not be great (or who are being aggressive with their rev rec)

Below is a high-level summary of the revenue to cash reconciliation I do (sometimes called a “cash proof”).

In theory, revenue should match cash after adjusting for related balance sheet changes that impact timing differences between cash and revenue - upfront payments (deferred revenue), pending cash collections (AR), etc.

If only it were that easy…because you always have to watch out for the “other adjustments” bucket. This is often the dumping ground for everything that doesn’t turn into cash. The biggest item is usually bad debt expense. Pay attention to growing/large bad debt expense

Here are a few other common things to watch out for when building this bridge:

• Sales Tax: You need to remove sales tax collected from your total billings number and your customer collections per the bank. It’s not your money…

• Gross vs. net presentation mismatches: An example of this is when a payment processor like Stripe nets fees before remitting. Your bank shows net cash but your revenue was gross.

• Foreign currency translation adjustments: When multi-currency lines are converted at different rates than what was booked you will have a difference if there are material changes.

Final Thoughts

These reconciliations matter because they force you to stay honest in what ARR actually is. If ARR doesn’t ultimately lead to revenue and cash then it’s a fake metric anyways.

Board Reporting. I have started to include an ARR to rev rec reconciliation in the appendix. It keeps everyone honest about what ARR is and the board loves it.

Revisit ARR Definitions: Most companies need to redefine / evolve how they define ARR for the AI era. Your pricing, contracts, etc., have changed so how you define ARR may need to change too. “ARR” should turn into GAAP revenue and it shouldn’t take a really long time for that to happen.

Can AI Help Reconcile? Obviously. And you don’t need some special AI prompt. You can probably just grab the link to this newsletter and feed it to Claude/ChatGPT and ask it to help you build the reconciliations. Then just build a Skill to do it going forward after you figure out all the buckets you want to see in the recon.

Do these reconciliations and explain them to your team and board. That builds trust and gets everyone aligned.

Footnotes:

• Join Rillion’s webinar next month and hear how to build your finance function for the AI era

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Nothing here is investment, legal, or tax advice.

I’ve done a lot of international hiring, and it gets complicated fast…

• Do we hire as contractors, employer of record (EOR), or set up an entity?

• How do we handle compliance risk and all the potential tax issues?

• What is this actually going to cost us? This year and in the future

Remote simplifies global hiring with the experts and solutions you need. Fewer surprises, better spend visibility, and less risk.

How To Hire Globally

The Cognition vs Factory Battle

Last week I wrote about the Legal AI drama between Harvey and Legora, but that is nothing compared to the latest Twitter/X fight amongst competitors…

A fight on Twitter between Factory ($5B valuation) and Cognition ($48B valuation) started after Factory’s CEO publicly said he fired Chris Degnan for unethical conduct.

It was a long tweet, but he definitely had an excellent hook :)

Chris spent 11 years as Snowflake’s top sales leader (eventually with the CRO title). He stepped down as Snowflake’s CRO in March 2025 and has been advising a bunch of companies since. He has been a GTM advisor and board observer at Factory, a major competitor to Cognition, for over a year.

So what was Chris’ alleged “unethical conduct”?

Factory CEO Matan Grinberg said the following:

For weeks, while he sat in our board meetings and advised our leadership team, he was also confiding with executives of our largest competitor.

The allegation was that Chris was advising Factory on strategic decisions, attending board discussions, accessing confidential board materials, etc., while also talking with their largest competitor about a CRO role. And that doing both creates an obvious conflict of interest and breach of trust.

Chris denies any wrongdoing

Chris responded to Matan with the below. Implying that there was no wrongdoing.

Conflicts of Interest

People go to competitors all the time. What’s the big deal?

It’s true, especially with sales reps. They are interviewing while still employed at a competitor. It’s pretty common. You hate to see your people go to competitors, but it is what it is.

The problem in this situation is the position Chris was in. He wasn’t just an employee. He was a GTM Advisor and Board Observer so he had access to confidential and strategic information. The argument is that because of those positions, Chris had a “fiduciary duty” to Factory that should have stopped him from having any conversations with Cognition while in those roles (more below on why the “fiduciary” part isn’t exactly true).

Keith Rabois (GP at Khosla Ventures) is an investor in Factory and he said the following:

I don’t disagree with Keith and it sounds like Chris may not disagree either

Chris refuted this implication by saying in his tweet that the last Factory board meeting he attended was weeks before the Cognition discussions started. But Factory also claims he had access to confidential board-level information and participated in strategic conversations while interviewing with Cognition, which Factory also argues was unethical.

Board Member vs Board Observer

Chris was a “board observer”, not a board member. The difference is important.

Like the name states, board observers mainly just “observe”.

• They don’t get a vote in board matters (like M&A approval)

• Aren’t technically held to “fiduciary duty” standards like board members

• Have limited liability exposure (because of the above 2 points)

• Don’t sit on committees (like audit or compensation committees)

• My general rule for observers in board meetings - “Shouldn’t speak unless spoken to”. Potential exception is if they are an expert on a topic you are discussing.

• Observers often just call into the meeting if it’s in-person

Observers are not “fiduciaries” with the same responsibilities as directors, but that doesn’t eliminate ethical responsibilities or an observer’s contractual confidentiality obligations.

There are typically two board meeting sessions:

1. General: Directors, observers, and often most of the exec team.

2. Closed: Just the directors, which includes the CEO. Sometimes the CFO for part of it (approval of financial plan, option grants, etc)

3. Optional: Board of Directors (excluded CEO): And then sometimes the board will kick out the CEO after the closed session so they can talk openly about management and how things are going.

Chris would likely have only been attending the open session because of his non-director status so he would have had access to most but potentially not all board materials and conversations.

What’s The Point of a Board Observer?

If they don’t get to vote and aren’t supposed to really say anything, what’s the point of a board observer?

It’s for people who don’t make the cut of being a board member (or they don’t want a director seat), but they have some leverage to negotiate an observer seat. Board member seats are typically reserved for:

• Founder/s of the company

• VCs leading fundraising rounds

• Independent directors (usually only appointed after the company is larger)

Common ways someone gets a board observer seat:

• Strategic investor: Often they don’t want a board seat because of potential conflict of interest problems that may arise in the future and/or to reduce liability exposure from being a Director

• Lead VC: Negotiates observer rights when it doesn’t receive a voting board seat. Not every lead VC will get a board seat (especially when there are many rounds of financing)

• Lenders: Large lenders may negotiate an observer seat in financing agreements

• 2nd person from lead investor: Sometimes a more junior person from a big investor will get a board observer seat so they can attend and learn

In our Factory vs Cognition story, Chris was a partner at RPT Partners, which invested in Factory so he likely got his observer seat from that, but Chris was also a GTM advisor for Factory.

VC Conflicts of Interest

Both Factory and Cognition had VCs coming to their defense. Two of the most prominent voices (Keith Rabois defending Factory and Vinod Khosla defending Cognition) both work for the same VC firm - Khosla Ventures.

Doesn’t a VC firm that invests in competing companies create a conflict of interest?

Yes, obviously investing in direct competitors can create conflicts of interest. The competing companies would be presented in portfolio review meetings, everyone could see the latest valuations, potential new fundraising participation, etc.

It can be really hard to be a good fiduciary if your VC firm invested in two very competitive companies like Cognition and Factory.

Keith said Khosla Ventures’ general practice is to only take a board seat at one company if they are invested in direct competitors. And in this case that is Factory.

This reduces some conflict, but still doesn’t eliminate it. The information sharing/visibility within the firm is still there for most VC firms. If the VC firm only has one board seat between the competing investments, that potentially means that the company with a director gets an unfair advantage from that director’s access to competitor intel that may circulate at the firm. I actually don’t think Keith or Vinod would purposely cross that line, but it’s hard to completely ignore what you hear about your biggest competitor when you hear/see stuff based on your VC firm’s relationship.

Many VCs have this problem. Investing in direct competitors on purpose is more rare, but many early VC portfolio companies pivot or expand to be highly competitive with other portfolio companies. When this happens, sometimes the appropriate answer is for one of the directors to step down.

Did Chris or the VCs act unethically?

If Chris was a “board member” (which includes fiduciary duties) and he was attending meetings and strategic discussions while actively interviewing at Cognition, it probably would be unethical.

But Chris was a “board observer” and he says he didn’t attend any board meetings when discussions with Cognition started and that he would never share confidential information. But…even if Chris was just an observer, if in practice he exercised more involvement and acted like a director then he would have fiduciary duties under Delaware law which would make his situation very messy (not sure if that is true in this case tho).

I don’t think Chris was intentionally attempting some corporate espionage or to act unethically. Just like I don’t think Keith or Vinod are intentionally doing anything nefarious when they are both investing in competitors while at the same VC firm. But good intentions don’t create an information firewall…

Whether intentional or not, having confidential strategic information about a competitor is going to influence your decisions. Although, in theory, it’s easier for VCs in this situation to control the conflict of interest since they don’t have to be involved in strategic decision making at both companies. CROs can’t really do that because part of their job is being strategic.

The Front Page Newspaper Test

Neither company is coming out a winner from this Twitter fight.

But it is a good reminder that everything we say or do could be made public, which reminds me of the old “front-page newspaper” ethics test:

Would you be comfortable with your decision if it was printed on the front page of the newspaper? If your colleagues, your board, your mom, etc. all knew about it?

Just replace newspaper with Twitter/X in this situation since that seems to be where all this stuff surfaces now…

I don’t have all the facts and details of this situation so I am not going to try to say who is wrong or who is right. That’s not the point of this article. But if you find yourself with a potential conflict of interest, the most important thing to do is to disclose it early and then figure out how to handle it.

Footnotes:

• Remote can help you hire globally while remaining compliant and providing transparent pricing (today’s sponsor)

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Nothing here is investment, legal, or tax advice.

Global hiring gets complicated fast when Finance can’t clearly answer two questions: What will this cost us, and who owns the compliance risk?

• With Remote, pricing is transparent and predictable, with no hidden fees.

• With Remote, uncapped indemnification gives stronger protection against compliance risk.

Fewer surprises. Clearer ownership. Better visibility into global employment spend and risk.

How To Hire Globally

The Harvey vs Legora Battle

I love a good public rivalry with juicy social media trolling. The Harvey-Legora battle is the Kendrick-Drake feud for those who know what “Pari Passu” means.

Harvey and Legora are both only a few years old and have seen epic revenue growth, which has led to skyrocketing valuations and large consecutive fundraises.

Both companies were among the very fastest application layer AI companies to reach $100M (and $200M) ARR. Legora went from $1M → $100M ARR in 18 months and then $100M → $200M ARR in 6 months…

Both have raised tons of money to fuel rapid growth, and both are burning cash to sustain that growth (more on this below).

But these two companies have many doubters about their long-term future. The comments are valid concerns:

1. They are just thin AI wrappers

2. They have negative/low gross margins so they will never be profitable

3. The products are worse than just using Claude/ChatGPT

4. How will they actually differentiate? The frontier models are entering their space

Is Eric (founder of one of the fastest growing law firms) right?

Do Gross Margins Matter?

Let’s start with the gross margin point because it is a key financial metric underlying the broader discussion of these companies.

AI products have significantly lower gross margins than their legacy SaaS friends, but they appear to be improving!

An issue though is how much gross margins can differ between AI products/companies. The range is so much wider than with legacy SaaS. Many even have negative gross margins.

Harvey and Legora have both been called out for weak gross margins:

Harvey’s Terrible Gross Margins

Harvey’s gross margins reportedly fell to -50% in June because:

1. Harvey primarily has flat seat-based pricing (doesn’t change with AI usage)

2. Underlying frontier models have been getting exponentially better, which drives more usage since it’s more useful (especially since all additional usage is free to customers)

3. Harvey released better AI capabilities earlier this year that drove more usage

Keep reading to see how Harvey flipped to positive gross margins within one quarter (50+ percentage-point increase 🤯)

Legora “Positive” Gross Margins

The CFO of Legora said the following about their gross margins in a post about the most important metrics for AI companies.

Having “positive gross margins” is better than negative but also a pretty low bar. A company with 0% long-term gross margins is worth $0.

I assume the gross margins are pretty low because their CFO specifically quantified other metrics (300% NRR and 95% GRR) while just saying gross margins are positive. Barely positive gross margins are also not sustainable given all the OpEx spend these AI companies have. So improvements still must come.

The second part of this Tweet reads like a clear jab at Harvey’s period of negative gross margins :) - “There’s a shorter route: Price below what it costs to serve”.

A Critic’s View of Negative Gross Margins

The skeptics will say:

Of course you can have explosive revenue growth if you are selling $1s for $0.50

There is unlimited demand for compute and frontier intelligence so if someone is willing to give you that same intelligence for a 50% discount then you will have almost unlimited demand. No brainer, right?

But what happens when gross margins improve and approach 0%? Probably no change. If there is unlimited demand for the underlying intelligence that you are reselling, then there will be near unlimited buyers for your thin AI wrapper at any discount.

But a truly thin AI wrapper will lose almost all demand once gross margins start to become positive because users will just switch to the underlying AI models since using the AI models directly would presumably be cheaper at that point.

A Positive View of Negative Gross Margins

What if temporarily negative (or very low) gross margins can create a longer-term strategic advantage that enables high gross margins?

This is obviously the bet of high-growth negative gross margin companies.

Here are a few arguments:

1. S&M driven - Drive initial customer adoption into a free/cheap tier to eventually get high overall gross margins by pushing the biggest customers into a profitable tier.

2. Build Adoption/Distribution - Get the adoption, get more data, and continue to add value on top of what the AI models can provide. Vertical software (like Legal AI) is a great place to do that.

3. Product research - Give away frontier-level intelligence so the product is great while you figure out how the product is used to make improvements while bringing down costs (fine-tune your own model, improve harness, or model routing, etc)

4. Falling Inference Costs - Inference costs are falling fast enough to underwrite current gross margin losses (costs have come down 10-100x in the last couple of years)

#4 is a bad reason to explain bad current gross margins. Yes, inference costs are falling but the AI model vs Legal AI (or a “thin AI wrapper”) decision is relative to the AI model costs… If AI model costs come down 100x then customers are going to just use the AI models directly if you remain a thin AI wrapper.

Improving AI Gross Margins

Companies have three options to improve gross margins based on their current product offering:

1. Change the pricing model

2. Solve the LLM cost issue

3. Cap AI usage so costs don’t exceed fixed pricing

Longer-term, you really need gross margins improving from your product differentiation and the material value you are adding beyond the LLMs.

Let’s look at how Harvey and Legora are attacking their gross margin problem:

Harvey’s Gross Margin Fix

Harvey’s gross margins clearly needed a major fix after it went deeply negative.

Harvey is shifting more to a consumption-based model, but didn’t want to force customers on it too soon so they focused first on #2 (solving the LLM cost issue)

Harvey made several changes to reduce AI costs, including releasing their own model in August that was fine-tuned from an open-weight model (Kimi K3, from the Chinese lab Moonshot).

• Claims their post-trained model can deliver better answers than the frontier in many cases

• Still uses frontier models when needed

• Their post-trained AI model costs can be up to 1/10th the cost of using just the frontier models:

  • Kimi K3 is 40% cheaper

  • Harness improvements

  • Orchestration improvements

If Harvey can offer their solution at a better price than frontier AI models (with good gross margins) and with equal or better responses for legal teams then of course every legal team will prefer their solution.

Legora’s Gross Margin Improvements

Legora’s “positive gross margins” are better but also not sustainable if they are really low, which they presumably are. They still have to pay for all the costs in R&D, S&M, and G&A….

Legora chose a combination of #1 and #2 (consumption pricing and solving LLM costs)

Legora introduced a consumption-based pricing model a couple of months ago for certain customer segments, so pricing will move more in line with customer usage.

Legora isn’t currently building their own model like Harvey. They are focused on some post-training for specific tasks, improving their harness, better orchestration, and model routing.

The CTO of Legora wrote a post about how “The best model is all the models” - clearly trying to differentiate themselves. What he said about the vertical AI opportunity makes a lot of sense:

Vertical AI companies are uniquely positioned to map the nuanced model map and spot “model arbitrage” opportunities - places where changing models gives more performance, lower cost, lower latency or the gain of other desired properties - that aren’t immediately clear from a public leaderboard. They can do this because they have two very valuable things: in-house expertise and valuable real-world usage.

What AI Stuff Goes Into Gross Margins?

Gross margins represent the ceiling on profitability.

Gross margins show how profitable a company can be if there were no other expenses (stop developing products, stop fine-tuning AI models, no sales reps, AI automatically does finance, etc.).

Below is a list of the main expense categories that go into COGS (and therefore impact gross margins). I also broke out other AI-specific expenses that should be in other sections of the P&L to show what specifically shouldn’t be in COGS.

I know a lot of companies that are getting the P&L classification of these AI costs wrong. Some common ones I have seen are:

• Fail to allocate AI model subscriptions (and dump it all to G&A)

• Free trials aren’t separated so it all goes into COGS. Note - not everything given away for free is S&M. Talk to your accountants.

• Accounting team doesn’t understand what vector databases are (or other data processing, retrieval or agent execution tools) so it all goes to R&D

• Some R&D costs may need to be capitalized under GAAP and then they get amortized into COGS. Yes, accounting can be confusing…

Final Thoughts

I will leave you with three things to think about:

1. Many generational AI companies are being built, but I agree with Harry that this feels like 2021 on steroids. Many financial metrics are being ignored in the pursuit of revenue growth. The thing is…the good companies aren’t ignoring them. They are very much paying attention to them, but they are temporarily sacrificing some metrics to build a long-term advantage. But many simply have terrible unit economics.

2. Gross margin reporting is important. If you aren’t properly tracking gross margins, your company will fail to make the right long-term decisions. Granular reporting is important - down to the product and even feature level

3. Thin AI Wrapper Test: What is the correlation between rising gross margins and sales demand? If demand falls off a cliff as gross margins go from 0% → 40% then it’s probably still a thin AI wrapper.

The question is whether today’s losses are building tomorrow’s advantage…or if it’s a continuous VC fundraising treadmill that likely ends in valuation destruction and tears.

Footnotes:

• Remote can help you hire globally while remaining compliant and providing transparent pricing (today’s sponsor)

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only.

Bonus:

Saw the below comparison between Harvey and Legora so sharing if helpful. Not sure how the survey was done (and sample size is pretty small), but still interesting. Obviously leans toward Harvey but Legora is very close on all of these metrics.

Every board is asking how the company is leveraging AI. At first they just wanted to make sure the company was using AI, but now they want to know:

• What are you doing that actually adds value for customers?

• What are the best use cases for internal work?

• Are you moving faster and/or more efficiently yet?

Brex’s The CFO Guide to AI Strategy is a great resource for CFOs to see where other finance teams are finding the best AI use cases right now. I have implemented most of this myself…

Download the Guide

Did We Pay a Fraudulent Invoice?

A few years ago, I woke up to a dozen texts from my AP team and Controller in absolute panic. Never a good sign…

We think there was fraud in a recent wire we sent

I have had a couple of scares like this in my career. Luckily, none of them resulted in a loss of money but a couple of them revealed some gaps in our controls and employee training.

I now walk through all our processes that touch cash and ask: “If I wanted to commit fraud, how would I do it?”

T&E Reimbursement Fraud

T&E fraud is one of the most common because every employee has access to do it, it’s harder to catch, and often employees can claim it was an accident.

The most common ones in my experience are:

• Personal expenses: non-work meals, travel, gifts, entertainment, etc., submitted for T&E reimbursements.

• Duplicate claims: same expense submitted more than once (same trip, same receipt, or same charge on card plus out-of-pocket).

• Unused bookings: flights, rooms, or events that were reimbursed before they were cancelled. For example, an employee is reimbursed but hides the airline credit to use for their next vacation to Tahiti.

Almost every company with any decent employee count has these frauds (or undetected errors). Sometimes these are legitimate accidents, but I know a good number of these are fraud…

AI-Generated Fake Receipts

Back in my day…there were limited (or sketchy) options to submit a fabricated receipt. But now? Thanks to AI, you can make a VERY convincing receipt almost instantly for anything you want.

The fraud methods are the same, but AI has removed the friction of the evidence layer.

Here is a fake receipt I created in ChatGPT. The receipt on the left was just a quick attempt and then I asked ChatGPT to keep iterating and find what made the receipt look like it was AI-generated and correct it and keep doing that until it can’t tell it’s AI anymore. It’s pretty good…

I ran the final fake receipt above through Claude and it concluded the receipt seemed real…Maybe if someone sits and stares at it there are a few potential clues, but if Claude thinks it’s real then I don’t think an accountant who is reviewing 200 reimbursements every week will catch it.

I probably wouldn’t risk my job (or soul) for $48 though. If you get caught submitting an AI-generated receipt then there is no excuse. But what if you are in IT or marketing and frequently put large expenses on your personal card for reimbursement? Maybe a few thousand dollars is worth it to some people?

AI-Powered Fraud in Accounts Payable

Employee expense reimbursements are generally small. The risk of T&E fraud is higher but materiality is usually pretty low. But the potential loss in AP fraud can be extremely high.

The Fraud Attempt That Almost Got This CFO

A couple of months ago my friend’s AP team almost fell for a pretty sophisticated attack:

They received a message from their vendor’s 3rd-party accounting firm that handled their vendor’s accounts receivable. The bookkeeping firm informed them that the vendor’s banking information had changed. The bookkeeping firm sent them a bank letter with the change, forwarded the email history with the vendor’s approval, cc’d two people from the vendor, and provided a phone number of one of the individuals from the vendor if they wanted to call and verify.

It all looked very legit.

• The bookkeeping firm domain had a legit-looking, professional website

• The bank letter had all the right things

• The forwarded email was well written with company-specific context

• The individuals that were cc’d from the actual vendor had the right domain (it wasn’t a fake domain)

Why would they assume it’s a fake request?

Well…one of their most important controls in AP is to verify any changes with a known/trusted contact. So when the AP team replied and added AR@[vendor].com for approval the fraud attempt was blown up.

Here is what the fraudsters did:

• Created a fake bookkeeping website and domain (super easy with AI)

• Likely used AI to create a fake email history that all seemed legit (no broken English and it had company-specific context)

• Used AI to fake bank and other documents

• Added fake vendor email addresses to make it look real. This part was actually really clever. The “@[company name]” was legit and they used names of real people on the vendor’s AR team. It was the vendor’s correct domain name. That built trust since the vendor was cc’d on the email. But while the domain of the email was accurate, the first part of the email was the fake part (so the email didn’t actually go to anyone at the vendor). For example, they used [firstname].[lastname] convention when the actual vendor used [first initial][last name]. They just wanted it to add legitimacy to the forwarded email.

• Fake call verification - while they didn’t call the number in the forwarded email, I bet the fraudsters were ready either with an AI voice or someone with perfect English

Controls to Prevent AI-Powered Fraud

Below are a few controls every team that touches company cash should consider. Not an exhaustive list.

• Banking Controls. I am continually shocked by the number of companies that don’t have appropriate banking controls in place. A VC wires them $25M and they just wing it…

  • ACH Debit Filters/Blocks - Prevent or restrict unauthorized ACH debits against your account by blocking them or allowing only approved originators.

  • Dual Administration - Enforces security controls by requiring setup and modifications to be approved by a second administrator. This mitigates the risk of internal fraud.

  • ACH/Wire Approvals - Set this up immediately. You should have all wires that are initiated by one person require approval by another person. You can also add a 2nd approver for wires over a certain limit (e.g. CEO approves everything over $100K).

  • Alerts: Banks offer a range of alerting options to help companies detect fraud.

  • Positive Pay - This is a control that provides detection of fraudulent, altered, or counterfeit checks through daily verification of checks presented for payment against a company’s check register.

  • 2FA (2-factor authentication): Please please set this up for both your company bank account and your personal account. It’s easy and adds a lot more protection.

  • Training: Make sure that the folks who initiate or approve wires have received training on best practices and how to prevent fraud. Any “urgent” wire request from the CEO is extremely suspect.

• Culture of Skepticism: Make sure the team knows it’s encouraged to question anything that seems odd or suspicious. AI-powered attacks are getting really good

• Segregation of Duties: The same person should not be able to create/change vendor banking information and release the resulting payment. Important to make sure no fraud is happening internally.

• Pay by Card: If you can, pay by card. It’s easier to dispute and card companies often have better fraud controls. Once an ACH or wire is sent, getting the money back is very hard (ACHs can be a bit easier though if you act fast).

• Verify and Establish Trusted Vendor Contacts: Get explicit approval from the trusted contact that was previously verified. Verify bank details or changes independently. It’s always worth the small extra friction.

• Bank-Change Reapproval: Any bank information changes should go through the same approval process as the first time (bank letter, verification, etc).

• Authorization and Approval Matrix: Define who can approve purchases, new vendors, vendor bank changes, and payment releases. And what the thresholds are for each. The payment approver should review the actual beneficiary and amount. Changes after approval should require reapproval. No exceptions.

• Spend Management Software: Brex (today’s sponsor), for example, has robust controls to help prevent and detect fraudulent/erroneous spend. Brex has seen everything so they built a simulated company full of expense abuse to train and test their AI audit agent.

Your corporate insurance won’t necessarily save you if you get duped. Two potential reasons why:

1. Insurance limits: A lot of insurance companies see increased risk here so the retention (aka deductible) is quite high and the maximum coverage per incident/aggregate is not that high. So one big fraud attempt may not be fully covered.

2. Callback verification condition: Even if you do have strong insurance, they can deny the claim if AP doesn’t independently verify payment instructions (such as calling a known/trusted number). Check your insurance policy and make sure your controls meet your insurance carrier’s requirements…

Scared Yet?

AI has removed the friction on the evidence layer for committing fraud. This is making fraud attempts really good. Your controls and systems need to keep up with the fraudsters.

We are going to see more companies fall for these frauds in the coming months. AI hasn’t changed the fraud methods, but it has made them much more believable.

Stay safe out there :)

Footnotes:

• Talk to my friends at Brex for a true AI-native spend platform

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Does not constitute investment advice. The author may own positions in companies discussed. Do your own research. Don’t make investment decisions based on anon online characters please

Every board is asking how the company is leveraging AI. At first they just wanted to make sure the company was using AI, but now they want to know:

• What are you doing that actually adds value for customers?

• What are the best use cases for internal work?

• Are you moving faster and/or more efficiently yet?

Brex’s The CFO Guide to AI Strategy is a great resource for CFOs to see where other finance teams are finding the best AI use cases right now. I have implemented most of this myself…

Download the Guide

ICONIQ’s Pacesetter Index Is Wild…

I have never seen higher revenue growth expectations than we have today.

ICONIQ released a benchmark report this week that showcases their “Pacesetter Index”, which is their top-quartile AI-native companies. If you want to be humbled about your metrics, then check out the below table…

A lot of VCs have companies that fit the “Pacesetter” profile in their portfolio right now, so it has shifted their expectations.

Report Caveat: If you just skim the report you will read “top-quartile” and think these are the metrics of the top 25% of companies. “Dang! We must be a bottom-quartile company because these metrics are crazy!”

Not really…Understanding firmographics and the benchmark context is important:

1. The Pacesetter Index is primarily ICONIQ portfolio companies. ICONIQ is one of the best VC firms so they get into the very hottest companies

2. The companies in the Pacesetter Index are hand-selected as “AI-Native”. These types of companies are generally growing the most right now

So when you read “top-quartile” in this report, you should think top-decile (or probably even higher) given it’s the top-quartile of AI winners at one of the best VC firms.

But it’s still valuable to see the metrics of these top companies because investors are comparing you to them.

VC: Grow Faster or You're Dead

Several VCs have gotten hate for saying something like Hemant Taneja (CEO of General Catalyst) - the gold standard of revenue growth from a few years ago (T2D3) isn’t good enough anymore.

Lots of people were quick to argue that T2D3 is still incredible and VCs are out of touch. “You can build a huge business with T2D3! The best software companies have been doing it for the past decade.”

Look…I get it, but Hemant is not wrong. His point is that venture dollars get allocated on a relative basis. So when the top companies are materially better than T2D3, then VC dollars will go to them.

T2D3 might become relevant again if these hot AI companies come back to earth after the initial rocket ship growth, but for now…VCs are expecting faster growth.

And that is what ICONIQ’s Pacesetter Index is showing us. These aren’t the only companies getting funded today, but the expectation is that something like these metrics are possible in order to get funding (and receive a valuation premium).

The chart below shows the time it takes these top companies to reach $100M ARR - we are now measuring it in quarters 🤯

Jason believes a similar timeline is required in order to raise venture capital today - you have to be able to go from $1M to $100M in <5 years (ideally in less time)

Public Markets: It Pays to Grow

And in the public markets we are seeing the largest spread between high-growth (>20%) and slower growth (<20%) revenue multiples.

Currently, high-growth companies trade at 3.4x the revenue multiple of a slower-growth company. Everyone wants high revenue growth because you get a much higher multiple since outsized growth is viewed as evidence that you’re an AI winner.

We are not “growth-at-all-costs” levels, but…it’s definitely being valued more today than the last few years…the relative importance of FCF margins has been cut in half over the past 3 years.

Below is an interesting chart that maps segments by AI model releases and how major releases can swing software stocks a lot. Infra & security remain the hottest with vertical software holding up pretty strong too.

Burn Baby Burn

Pacesetters (ie super fast growers) are unsurprisingly burning A LOT of money in the <$100M segment. And that shifts a lot once they are >$100M.

But the sample sizes on both of these are really small for the Pacesetters, so take it with a grain of salt.

Benchmarks In The Wrong Hands

Too many folks benchmark their revenue growth to companies at a very different scale or with a very different profile.

We have top-quartile growth because the median public company revenue growth rate is 12% while ours is 18%!

Two things wrong with this:

1. You are at $50M of ARR so you can’t compare your growth rate to a company at $800M of ARR

2. That median 12% revenue growth also has a median 20% FCF margin. Yours is -10%

I am constantly surprised by the number of CEOs (and finance leaders) who make this mistake. Or they cherry-pick metrics without understanding/showing the full picture. They present it to the board like they are doing awesome. Board walks away thinking you are stupid…

Benchmarks are interesting, but be careful how you use them.

Final Thoughts

VCs are obsessed with revenue growth (part of their job description). If there is a new cohort of companies that are growing materially faster, that’s who they will wire their money to.

Many VCs will get burned badly on a few of these Pacesetter investments because the durability of the growth will turn out to be very weak in the long term. Revenue growth durability is one of the most important drivers of long-term valuation, but it’s really hard to predict, so investors often over-index on current growth. Grow fast, but make sure you can build a durable revenue engine.

There are 2 paths to raising money from VCs today. Many VCs are narrowly focused on #1 right now, but folks like Gokul are also looking for companies that might be growing slightly slower but are efficient and can have very durable growth.

1. Get Pacesetter-level metrics: Super fast revenue growth and likely high burn

2. Fast Growth & Efficient: Growth is strong, but not “Pacesetter” levels (more like T2D3). And it’s highly efficient

And if you can’t do #1 or #2, then you at least need a plan to get to FCF breakeven so you can control your destiny. But if you want a decent valuation for an eventual exit, then you still need to figure out how to grow faster.

Footnotes:

• Check out The CFO Guide to AI Strategy and figure out how your team should be leveraging AI to get the most ROI

• Full Report: ICONIQ - State of Scaling 2026

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Not tax, legal, or investment advice. The author may own positions in companies discussed. Do your own research.

In the age of AI, finance teams need to move fast while staying in control.

Brex is the intelligent finance platform that gives you the AI-native workflows to move fast (handling receipts, memos, transaction monitoring, policy enforcement, etc) while staying in control and eliminating manual busywork.

Join companies like Anthropic and Zoom that are moving faster because of Brex.

Learn More

Pulley (Carta Competitor) is Shutting Down

The founder of Pulley (cap table software) announced this week that they are “winding down operations”. No PE bailout, no fire sale acquisition to a competitor, Bending Spoons isn’t coming to the rescue, and not even an asset acquisition?

The biggest question on social media was “Why couldn’t Pulley get acquired?”. They seemed like a strong competitor to the 800-pound gorilla (Carta). They raised $50M from Founders Fund and Stripe. And had a decent amount of ARR - rumored to be ~$20M.

So what the heck happened?

The LinkedIn message from Yin Wu (Pulley founder) said a few interesting things:

1. They have been working with Carta for the “last few months” on the transition

2. Carta is honoring all Pulley contract terms

3. Implied that the majority (but not all team members) are gone and that she (and others?) will be building something new

Given the way she praised Carta and talked about how great they are, I think it’s safe to assume that they aren’t just “shutting down” with nothing. There is some sort of deal/incentive with Carta (more on this below) but also not really an acquisition.

So what happened to Pulley?

Obviously things didn’t go as planned…

I have been part of transactions like this from different sides, so let’s walk through how things led up to Pulley “winding down”, Pulley’s rumored deal with Carta, and why VCs are incentivized to choose speed over highest-value deal.

Revenue Growth Stalls:

Pulley raised a decent amount of money ($50M) so their mandate was to grow fast. Revenue growth obviously slowed significantly since their $40M Series B in 2022.

And when revenue growth is slow (like <30%) at anything near Pulley’s small ARR scale, it won’t be a venture scale outcome. So the VCs either want 1) a path to massive reacceleration or 2) get their money back as soon as possible.

I am going to guess that Pulley’s revenue growth has been slow for a couple of years. When they raised their Series B they probably had $5-10M of ARR and after ~4 years they are only at ~$20M. That’s way too slow for a VC-backed company at their last valuation - the exit waterfall math was broken.

The Search:

Below is typically what happens in a situation like Pulley:

1. Growth slows below an acceptable venture scale outcome

2. Management team tries different things to reaccelerate

3. #2 doesn’t work and they likely burn even more cash trying

4. Founders get tired and/or board doesn’t think they can turn the ship around. They suggest running an M&A process

5. Board and founder have quiet conversations with most likely acquirers

6. If #5 doesn’t work (or they do it in parallel), M&A advisors get hired to run a more formal process

At this point, everyone knows it’s not going to be a great outcome. VCs are ready to just get some (hopefully all) of their money back and move on. Founder is tired and doesn’t see a path toward a good outcome.

While M&A advisors are hired, a lot of these deals get done through the VCs network (like another portfolio company).

In Pulley’s case, Carta was clearly at the top of the list for a soft landing.

The Deal:

Pulley’s rumored $20M of ARR is definitely worth something so of course they aren’t just shutting down and telling their customers to move over to Carta out of the kindness of their hearts. Many believed that to be the case when they read the announcement…

According to Jesse Tinsley (CEO of MainStreet) who heard it from multiple sources, Pulley took an earn-out deal with Carta that might return up to 50% of the pref stack, which is $50M (so earn-out of ~$25M).

What’s an earn-out?

An earn-out is cash paid after closing that has to be “earned”. The seller “earns” the payments over time if the acquired business hits agreed-upon targets after the deal closes.

In Pulley’s case, this “earn-out” feels more like a commission. I am imagining something like “For every Pulley customer you get to migrate to Carta you get x% of the ARR”. Which would explain why they have been talking to Carta for months and the founder’s LinkedIn post was very complimentary of Carta and pushing everyone to move to Carta.

What about cash on the balance sheet?

The cash remaining after paying all liabilities and wind-down costs also gets distributed to shareholders.

So let’s say Pulley had $15M in the bank and got a $25M earn-out with Carta. That’s $40M of distributions. Not sure if this is accurate, but let’s assume it’s directionally right. This would mean that no one gets any money except for the investors since they have $50M worth of preferred shares.

We all know most startups fail, but always a tough pill to swallow when no common shareholders see a dime on exit.

So no one at Pulley will make any money?

Well…not necessarily. Often these deals come with carveouts or retention packages for a few key employees if they need some folks to stick around. Or in this case, encourage Pulley customers to migrate to Carta.

So founders often receive a small carve-out that’s separate from the deal consideration since all deal consideration gets eaten by the pref stack. The VCs want as much of their money back as possible so they want the founder to be motivated to help make that happen.

Could Pulley have gotten a better offer?

Maybe, but it’s hard to talk to every potential acquirer. And sometimes getting a deal done quickly is more important than a slightly higher acquisition price.

On Twitter/X, Jesse threw out a $70M acquisition offer to buy Pulley. But obviously this is non-binding since he doesn’t have any details of Pulley’s business.

Take this offer with a big grain of salt. Jesse doesn’t have all the context (ARR, metrics, issues, etc) and he knows it’s likely too late anyways. Some of this is Jesse just marketing so other companies considering shutting down in the future will go to him first. But…he has pulled something similar like this before when he acquired Bench…

It is odd that Pulley didn’t reach out to Jesse at all though. He is a serial acquirer and Pulley fits almost perfectly for what he looks for.

Having said that, there is definitely more to the Pulley story than we know yet. Maybe there was time pressure, or the Carta deal is better than we think, or Pulley is in worse shape than we think, or founder/board believed no offer would materially beat Carta’s.

The Dead Zone’s Impact on M&A

The Pulley team isn’t stupid. Neither is their board. If there is a “soft landing” or decent outcome then they will try to find it.

In general, everyone would love to find the best deal possible. But sometimes speed or high confidence in a deal closing becomes more important to the decision makers.

Not saying it was the case with Pulley, but folks should understand the “Dead Zones” in the exit waterfall where a VC’s payout doesn’t change as the valuation increases (even though others do benefit).

Pulley raised a total of $50M with the most recent round (Series B) of $40M at a $245M post-money valuation. For Founders Fund, there was no difference in their payout between a $50M acquisition and a $245M acquisition. Even the Series A wouldn’t see any benefit until it reached ~$71M. It sounds very unlikely that Pulley could have been acquired for more than $70M so investors were not financially motivated to optimize the deal for a few million dollars more. They are motivated by deal speed and making sure the deal gets done though…

Final Thoughts

There are a lot of founders getting tired right now. They raised and built in a pre-AI world. Things are a lot harder today and many want to build something else that they are excited about for a post-AI world.

I get it…building and selling cap table software probably feels A LOT different today than a few years ago.

On the other side though, this can be a big opportunity for many companies who are in a strong financial position. There are tons of Pulleys out there who are ready to move on.

Footnotes:

• Talk to my friends at Brex for a true AI-native spend platform

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Does not constitute investment advice. Certain ARR, cash-balance, earn-out and transaction-value figures discussed above are estimates or third-party claims and have not been independently confirmed.

Every board is asking how the company is leveraging AI. At first they just wanted to make sure the company was using AI, but now they want to know:

• What are you doing that actually adds value for customers?

• What are the best use-cases for internal work?

• Are you moving faster and/or more efficiently yet?

Brex’s The CFO Guide to AI Strategy is a great resource for CFOs to see where other finance teams are finding the best AI use-cases right now. I have implemented most of this myself…

Download the Guide

The Highest Paid Person

Many people assume the CEO, CFO, and other executives are the highest paid people at the company. A benefit of being in finance is you get to peek at everyone’s compensation, which sometimes comes with questioning your career choices.

There are two components of an employee’s total compensation:

1. Cash - base salary + bonus

2. Equity - stock options, RSUs, etc

For private companies (particularly those without an active secondary market), the highest cash-basis earners are almost never the exec team. If they are the highest paid, then something is probably wrong at the company (like flat revenue growth).

Who’s getting the most cash?

It should almost always be your sales reps. How many reps are earning more than the exec team is usually a function of company size and how many sales reps you have.

Top 10 Paid Employees (cash-basis) at ~$250M ARR Mid-Growth Company

I recently spoke with a CFO at a mid-size software company with decent revenue growth. Below were their top 10 highest paid folks:

Notice anything? It’s mostly sales folks…Other than the CRO (who owns sales), no other exec was in the top 5 (but they are doing OK too :)

How Much Does the Top Sales Rep Typically Make?

Below is survey data from RepVue on how much the top earner makes in a year (base, commissions, bonus, etc)

Most of these top-earning sales reps would be the highest paid people at their companies (with the obvious exception of the public ones).

Are we paying sales reps too much?

As painful as it might be to approve those 6-figure commission checks, you can’t conclude that just because a sales rep (or a few) are making a ton of money that the commission plan is broken.

All else being equal, I would love to hire great sales reps and pay them half as much and still close the same number of deals. Efficiency goes way up, right?!?!

Nah. That wouldn’t actually happen…

Your best sales reps are highly paid, but also they create the most sales efficiency. The reps you should be worried about are the mediocre ones. They are the expensive ones.

Example: Top Rep vs Mediocre Rep

The below compares two enterprise reps with the same OTE and quota. Except one rep closes $1M in ARR (or 77% attainment) while the other rep closes $5M (or 385% attainment). The “Top Rep” gets massive accelerator commission rates (20%) on everything over $1.3M in bookings while the “Mediocre Rep” earns just 11.5% commission rates on everything he closes.

CEO: Whoa…We are paying “Top Rep” too much! He is getting 20% commission rates and made over $1M.

Me: If you compare Top Rep’s cost to ARR, he is only 74% the cost of Mediocre Rep…He is our highest ROI Rep.

But even the above scenario doesn’t capture the true value difference based on my experience… Because the best reps usually:

1. Discount less - so gross margins are better on their deals

2. Qualify accounts better - so churn is lower on their deals

3. Set better expectations - Fewer concessions and less CSM work

Not saying they are perfect at all of these, but the mediocre/poor reps are usually more desperate and less experienced in all of these areas. If you haven’t, build a report to look at average discounting % by rep. It’s usually interesting…

How Much is Too Much?

While top-performing reps are more efficient, CROs and CFOs need to figure out the best balance - pushing higher revenue while improving GTM efficiency. If all your sales reps are crushing quota, then your comp plans aren’t aggressive enough.

If you average the attainment numbers of all the companies from the RepVue chart then you get an average of ~60% of reps hitting quota. Ie only 60% of sales reps are at least getting their OTE while the remaining 40% are not.

There are a few different thoughts on what attainment targets should look like:

• Dave Kellogg says 80% of reps at 80%+ attainment

• Others say 80% of reps at 100%

• Team-wide quota at 75%

• And many other versions of the above

Too high, not aggressive enough. Too low and the best reps will leave and a sales death spiral will happen. So if your top 10 earners are all sales reps then you may have a problem (not aggressive enough).

Two things are always true for growing companies:

1. Quotas are increasing

2. Territories are shrinking

What About Total Compensation?

Total theoretical compensation is almost always the highest for the executive team (CEO, CFO, CRO, COO, GC, CMO, etc) because of the size of their equity grants.

The top sales rep from my example above (making $870K) has VERY little equity. Like 0.005% of fully diluted. Maybe he will make $100K - $200K if things go really well with his equity.

The executives have 100x+ that…

If the company is public or has a decent exit, then the exec equity will be worth a lot more than the difference in cash-basis compensation. But…for most companies the “decent exit” is a BIG if.

Below is Fastly’s pay mix for its executives (CEO vs other execs). Only 11% of compensation is cash (salary and bonus)…The rest is tied to equity that is either service-based or performance-based.

Below is the pay of 5 public company CFOs and their relative cash versus equity mix.

Why I Love When Sales Reps Make a Ton of Money

And here is why every CFO, CEO, and any other exec should be cheering when their top sales reps are making a lot of money.

$1M incremental ARR booked at accelerator commission rates is $200K in commissions to the sales rep. But it also represents $100K in additional equity value for the CFO (assuming 1% ownership and a 10x revenue multiple)…And I am, theoretically, getting this much richer every time a deal is closed (assuming there is, or will be, a market for my shares).

We want sales reps worried about commissions, while execs worry about enterprise value.

Final Thoughts

Yes, sometimes I get jealous when I approve 6-figure commission checks but then I remember that hopefully means things are working well. And that my equity value is growing.

Occasionally an easy whale deal will fall into a sales rep’s lap, but the consistent top performers work REALLY hard and are 100% worth the massive commission payouts.

Footnotes:

• Check out The CFO Guide to AI Strategy and figure out how your team should be leveraging AI to get the most ROI

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Not tax, legal, or investments advice. The author may own positions in companies discussed. Do your own research.

In the age of AI, finance teams need to move fast while staying in control.

Brex is the intelligent finance platform that gives you the AI-native workflows to move fast (handling receipts, memos, transaction monitoring, policy enforcement, etc) while staying in control and eliminating manual busy work.

Join companies like Anthropic and Zoom that are moving faster because of Brex.

Learn More

The IPO window seems mostly closed except for the very best AI companies. Also, there are several public tech companies that IPO’d during the ZIRP era that in retrospect should never have gone public (AI accelerated some of their downfalls)…

This means that M&A has become an even more important source of liquidity and “soft landing” for companies.

One such public company is BigCommerce (recently changed name to Commerce.com):

• Stock is down 97% since the IPO

• Basically zero revenue growth

• Barely profitable (3% NTM FCF margins and near breakeven GAAP profitable)

What is a company like this worth?

Not much…BigCommerce trades at a 0.6x NTM revenue multiple, which is the lowest of any public software company that I track. Shopify, its closest public competitor, trades at 11.5x.

During peak software stupidity (2021), BigCommerce traded at 13.5x while Shopify traded at 37x. Relatively, BigCommerce has fallen A LOT more …

A Prediction with a Meme

As I was looking at BigCommerce and thinking “these guys definitely shouldn’t be a standalone public company anymore”, I posted the below meme.

Narrator: Bending Spoons (the serial SaaS acquirer) is creepily watching BigCommerce as a potential acquisition target

I have no idea if Bending Spoons is looking at BigCommerce (although at least a couple senior Bending Spoons folks liked the Tweet :). But…I do think it could be a decent acquisition for someone. For the right acquirer, they could probably pay up to 2x the current valuation and do well.

If you are not familiar with Bending Spoons’ game….they are a serial acquirer of digital/SaaS businesses and they IPO’d in July.

They buy struggling businesses and “transform” them into cash flow machines by cutting costs, maintaining/improving product, and usually raising product prices (by a lot sometimes).

Below are some recent Bending Spoons acquisitions:

Their business sounds a lot like private equity, but with one important difference - they never plan to sell (and never have) any of the businesses that they acquire. They don’t buy with the plan to sell in 3-5 years for 3x the purchase price like PE. They get their returns by turning acquisitions into cash cows that pay for themselves (rather than relying on a future exit)

It’s not just something cheap that they are looking for. Bending Spoons wants a product that is sticky

There are two good ways to look at this from a numbers perspective:

1. RPO (Remaining Performance Obligations)

RPO represents all contracted revenue that hasn’t been recognized. It’s different than deferred revenue (which is on the balance sheet) because it also includes unbilled but contracted amounts.

BigCommerce has $201M of locked in revenue, which is crazy compared to their current $220M enterprise value….They basically have locked in future revenue equal to the company’s valuation! *note - the actual amount might be lower because customers go bankrupt or some other extreme circumstances but GAAP requires RPO to be a pretty sure thing.

Some of this future revenue is already reflected in cash because it was paid upfront though. To calculate total contracted unbilled revenue you need to subtract deferred revenue from RPO.

Since there is only $70M in deferred revenue, there is $131M of contractual billings that an acquirer would be able to invoice and collect. Pretty good relative to their valuation!

2. NRR and GRR

PE companies and acquirers like Bending Spoons will spend a lot of time here. It usually won’t be too high otherwise the company will likely be too expensive, but if it’s too low then Bending Spoons would be concerned if the unit economics break and they can’t turn it into a cash machine.

The software business model is built on high NRR and GRR (although most public companies don’t report GRR so we rely on NRR)…

During the 2021 days, BigCommerce had 118% NRR and has dropped to ~95% over the last couple of years. But at least it looks steady! And slightly improving.

It’s not the most straightforward acquisition target for Bending Spoons, but I can definitely see a path for someone to make some money acquiring them.

BigCommerce Rejects Acquisition

A public company (Rezolve AI) actually tried to acquire BigCommerce earlier this year. Rezolve is an AI ecommerce platform so the interest made sense, but their bid apparently did not make sense to the BigCommerce board.

Rezolve’s acquisition attempt is actually pretty funny….

First Acquisition Attempt - Feb 2026

• Strategy: Rezolve went to BigCommerce’s board (non-hostile)

• Price: Rezolve offered 1 share for 1 BigCommerce share. The problem was Rezolve’s stock traded at $2.15 vs BigCommerce’s $3.02…a 29% discount to BigCommerce’s then-current price.

• BigCommerce Board: Rejected the offer asking why shareholders would want an acquisition at a huge discount to current prices.

What’s funny is how Rezolve responded and what they did next:

Rezolve’s Response to Rejection

Rezolve didn’t hold back in their response to BigCommerce’s rejection.

My favorite line: ”Your board rejected a gift”….lol

Rezolve argued that BigCommerce is a “zombie stock” and that the “thinly traded screen price” (stock price) was fake because the volume was so low.

Rezolve went on to say that one Wall Street analyst valued their stock at $11 (4x what they were trading at) so BigCommerce was actually getting a really good deal 🤣. You don’t see that argument every day….

Side Note: My dad recently told me he strongly believes that the OnlyCFO Newsletter is worth $1,000,000,000 so I plan to use that valuation as acquisition currency to build my finance empire.

Second Acquisition Attempt - April 2026

Since BigCommerce rejected their acquisition offer, Rezolve decided to take the offer directly to the shareholders this time (getting hostile!). They wrote an open letter to the shareholders.

But…the offer price changed :)

Rezolve’s new offer: 1 Rezolve share for 2 BigCommerce shares (deal was cut in half and represented a ~50% discount to BigCommerce’s current stock price lol)

Daniel (Rezolve CEO) refers to his bid as a “rescue” 🤣

Rejection and Adopting a “Poison Pill”

You won’t be surprised to hear that BigCommerce also rejected the new offer.

But Rezolve’s aggressiveness prompted BigCommerce to adopt a “Poison Pill” to deter any hostile takeovers. This is common once hostile takeover attempts begin.

What is a “poison pill”?

It’s a defense that makes it much more expensive for someone to acquire a company without the board’s approval.

Here is how BigCommerce’s poison pill works:

1. New Right: Every shareholder gets a new right attached to their shares. The right allows BigCommerce shareholders to pay $13 for $26 worth of BigCommerce stock but only if certain triggers occur (see below)

2. 10% Threshold: Once a person/group acquires 10% ownership (20% for certain passive investors) then the new right is triggered.

3. Exclusion: The right is triggered for everyone except the acquirer (their right is voided). So everyone buys up a lot of stock at a discount and dilutes the potential acquirer.

In theory, an acquirer could just trigger the poison pill dilution and keep buying shares to acquire the company but that would be really expensive with a lot of unknowns. So no one actually does that.

What the hostile acquirer would do next is start a “proxy fight” to get shareholders to elect directors who support their offer and waive the poison pill. Rezolve kind of threatened a proxy fight but doesn’t seem to have actually pursued it, but they told shareholders to vote against the current board members that were up for election.

Why “Classified Boards” Matter

BigCommerce has a “classified board”, which means that there are different classes of directors that serve in staggered terms so that only a minority of directors are up for re-election in any given year.

If Rezolve attempted to continue with the hostile takeover, BigCommerce’s classified board would have also slowed them down because it would take a while for them to get board control.

What’s Next? Does Rezolve give up?

Last month, BigCommerce reported disappointing quarterly results and Rezolve made a press release that basically said “I told you so”. Rezolve didn’t provide a new acquisition offer but they suggested it would be even lower than before if they did…

BigCommerce’s stock fell ~38% over the next three days after their quarterly results. Maybe Rezolve was right?!?! Was the stock mispriced…?

If you track the two companies’ stock prices since the original offer (1:1 share exchange), Rezolve is up only 3% (guess the Wall Street analyst was just early :) while BigCommerce is down 15%. So far, good decision by BigCommerce…

When Will BigCommerce Sell?

I don’t know a ton about BigCommerce other than their financials seem pretty broken and it would likely be more valuable if they weren’t a standalone public company (acquired by Rezolve, Bending Spoons, etc)

But obviously price matters for current equity holders.

It’s hard to see a deal like this getting done for a price below their current stock price. So acquirers will need to either:

1. Wait for the stock price to continue to collapse (but how much further can it fall?)

2. Be willing to pay a 30%+ premium from today’s price

Based on BigCommerce’s performance and continued weak outlook, I wouldn’t be surprised if something happens soon. Seems unlikely the acquirer is Rezolve based on the history though…And it will likely need to be a non-hostile acquirer given the poison pill and classified board.

Maybe someone like Bending Spoons? Maybe…

Footnotes:

• Talk to my friends at Brex for a true AI-native spend platform

• Subscribe and share with your friends

Subscribe now

*Disclaimer: Informational and educational purposes only. Does not constitute investment advice. The author may own positions in companies discussed. Do your own research. Don’t make investment decisions based on anon online characters please

In the age of AI, finance teams need to move fast while staying in control.

Brex is the intelligent finance platform that gives you the AI-native workflows to move fast (handling receipts, memos, transaction monitoring, policy enforcement, etc) while staying in control and eliminating manual busy work.

Join companies like Anthropic and Zoom that are moving faster because of Brex.

Learn More

SaaS is Back Baby!!

Last week felt like a major vibe shift for the SaaS industry. Maybe SaaS isn’t dead after all? The SaaSpocalypse is over and investors just got it wrong…right?

It definitely looks that way based on public software companies’ stock performance last week (and over the past few months). Bessemer’s cloud index is up 55% in 6 months and it’s back to levels not seen since early 2022. SaaS has even outperformed the semiconductor industry over the last 6 months 🤯

Or…is this just a “dead cat bounce”?

Not all software companies are the same…

The SaaSpocalypse was never going to be evenly distributed. Investors treated all SaaS names like AI was going to send them all to zero by tomorrow and now we are doing the same thing on the way up (like they will all be fine).

I have gone through lots of SaaS renewals in 2026 so I put together the 10 biggest reasons why I have said “I am churning” this year. The companies that grow despite these threats (or even turn these threats into advantages) will be just fine.

10 Threats to SaaS Companies

1. Increasing AI Spend Squeezes the SaaS Budget

Every CFO is dealing with the same thing: If AI spend increases (which it probably has), then something else in the budget needs to decrease. Only exception is if you are accelerating revenue by a lot….then maybe you can have both.

AI may indeed reduce headcount, but you better believe that every CFO is pushing on existing SaaS spend as well.

Boards/investors won’t have much patience for higher spend because of AI but no change in revenue growth and/or efficiency…

2. The Value I Get From Claude Broke My Brain On ROI

You want me to pay 2x what we pay for Claude for your little workflow tool?

This makes everyone question how much they are paying for all their tools. It certainly has made me push hard on pricing (both new vendors and renewals). My bar for value relative to price has skyrocketed…

Related point below…And is the reason a lot of companies are starting to see huge price deltas. Are you selling to someone that is AI-pilled or an AI-noob?

3. Exponential Growth In Competition

AI has created an explosion in competition. Many of these competitors are new but AI has accelerated their products to get to feature parity really fast.

And there are two big things that can give them an advantage:

1. AI-native built so AI agents often work better with them

2. Built with less capital and more efficiently so they can be more price competitive

And threat #4 (below) has made the increased competition an even bigger problem for the incumbents.

4. Decline In Switching Costs

It’s becoming so much easier to churn. A few years ago I would have hated to churn off most finance tools because it was such a headache to switch vendors. Long implementation, costly consultants, lots of risk in moving, integrations, etc.

It’s so much easier today.

Most of us don’t have the patience for a long implementation anymore so it’s become a priority for vendors (if you have a long implementation you are ngmi).

5. Vendors Keep Expanding Into Each Other

Most of my vendors have massively accelerated and expanded their roadmap because of AI (as they should). This has meant that several of my vendors’ products are starting to overlap.

For example, one of my larger accounting vendors added a tool (for free) that I was previously paying $50K/year as a point solution…Was it as good? No…but it was good enough. Hard to justify $50K when I got something 80% as good for free.

AI is accelerating roadmaps and vendor surface area. The benefit will accrue to the stickier core tools.

6. Vibecoding

The mere threat of vibecoding or doing something natively in Claude is often enough to get your vendor scrambling to get you a “special deal”. I have played this trick a couple of times…

But sometimes it’s more than a threat. I am not on team “vibecode all your internal tools and therefore SaaS valuations are going to zero”. But…building internally can make sense when customization is important for you. The customization possibilities are one of the biggest reasons in my opinion to build internally.

Pro Tip: If you just want better pricing, then start the conversations early (4+ months before renewal). Then the vendor has plenty of time to panic, escalate, get pressure from the board for their high churn, etc so they make “special exceptions” on pricing.

7. New Pricing Model Is Stupid

I have seen a lot of pricing model changes in my vendor renewals this year. Everyone is trying to figure out how to price in an AI usage-based world.

It’s hard. There will be a lot of pricing experimentation.

But…on the buying side, my initial thought when I hear “We have a new pricing model” is “Great, how are you trying to screw me?”

The vendor’s job is to show:

• We are not trying to screw you (this year or next year as you scale)

• Here is how pricing works and why it makes sense (should be easy to understand)

• This is how you can forecast it

Fail at any of these (but especially the first one) then I am looking at alternatives.

8. AI Has Changed What I Value

AI has changed what we value from our software. For instance, UI matters a lot less today. I am logging into my finance tools less every day. I am relying more on AI agents or working directly in Claude to interact with my software tools.

Does your tool play nice with AI agents and Claude/ChatGPT?

I am probably going to churn from any tool that doesn’t have a prebuilt MCP today.

9. Vendor Not Accelerating Product Roadmap

I really don’t like taking demos. But now I tell my team that we need to occasionally take demos (especially before annual renewals) to see what else is out there.

It takes 30-60 minutes and helps us better understand what others are doing with AI. We churned from one legacy vendor after we discovered how much more we could do with a more AI-native one.

In the old world, SaaS renewals felt almost guaranteed because no one wanted to move. But that ARR feels a lot less “recurring” today because at every renewal I am considering other options.

10. AI Reduces Headcount/Seats

Obviously.

Everyone has heard enough about this so the only thing I will add is to make sure you don’t make stupid pricing changes (#7) because you are rushing to get out of the seat business.

Footnotes:

• Talk to my friends at Brex for a true AI-native spend platform

• Subscribe and share with your friends

Subscribe now

*Disclosure: Nothing discussed in this post is financial, tax, or investment advice.

Want to screw up a fundraising round, an acquisition, or an IPO?

Then don’t pay attention to indirect tax (sales tax, VAT, and GST). You will create a big liability that is hard to quantify without a ton of work. Ask me how I know…

It’s why one of the first things I do when starting a new job is 1) make sure we have good sales tax software and 2) a recent sales tax analysis has been performed.

With Numeral, you get both of these. That’s why Supabase, Character AI, and more than 3,000+ businesses have turned to Numeral’s global solution.

Check out Numeral

Joining [Hot Company] as an Advisor

I just signed my agreement to be an advisor at a pre-seed startup. I don’t normally take advisor roles because I stay pretty busy with my normal job, writing this newsletter, and spending time with my family (including my kids’ never-ending sporting events). But I made an exception for this company….

Will it make me rich? Or will it take a lot of my time with no treasure chest at the end of the rainbow?

Statistically, it is definitely the latter…

It’s the same as early-stage investing, but a much more concentrated bet since there is no way I can be an advisor to the same number of companies that early-stage investors bet on. Early-stage investing only works because of the power law…1-2 investments will return the entire fund and make up for the other 20+ losers that go to zero. This is why early-stage VC funds invest in ~20-30 companies.

So why be an advisor?

There are lots of good reasons to be an advisor:

• Smart team that you are excited to work with (most important)

• They are building in an interesting space

• Opportunity to learn new things

• Build your network and connect with new people

• If they are successful, could set the stage for you joining full-time

Unlike a job where money is a driving factor, don’t become an advisor unless you will learn and enjoy it. And maaaybe (if you are lucky) you might make some money.

The company I am advising checks all of these boxes for me, but I also believe it has the potential to be a massive success (incredible founders, tackling a big problem, they have deep expertise with the problem, and they have me as an advisor :).

Since advisor stuff is top-of-mind…there are a few topics I want to cover:

1. How much are advisors paid?

2. Is the dilution from advisors worth it? What do advisors even do?

3. How to structure an advisor agreement?

4. How do I become an advisor?

5. Bonus content: Did a Cursor advisor make $384M in ~4 years?

Subscribe now

Advisor Compensation & Dilution ROI

Hard Truth: 90%+ of advisors are not worth the dilution. Yes, I made up this stat but it’s definitely true based on my experience.

They take too much equity and provide little value. Common reasons include:

• Failure to set proper scope/expectations

• Misaligned compensation for scope

• Wrong person and/or experience

• The people and expertise you need can change quickly at early-stage companies.

Why don’t you just fire them as soon as you realize they aren’t providing value? They are on a vesting schedule so you can stop the dilution bleeding, right?

You would not be wrong for asking that question…but it rarely actually happens. I’ll explain why further below.

How much do advisors get paid?

Let’s start with the difference between an advisor and a consultant because that can create confusion:

• Advisor:

  • Compensation: Usually equity.

  • Role: Review decks, discuss strategy, provide product feedback, make intros, etc. Advisors provide their perspective and feedback on work that is already created.

• Consultant:

  • Compensation: Typically cash. But could be equity for cash-strapped startups.

  • Role: Produce work. Consultants create decks, perform analysis, write code, etc.

Carta has great benchmarks by fundraising stage that generally line up with my experience. But like most benchmarks…the “right” amount of equity for an advisor can vary a lot.

Equity percentages fall (just like employee equity packages) as a company scales because there is less risk and the equity is theoretically worth more. But also, as a company scales, the need for advisors decreases because a lot of the gaps that advisors fill are replaced by full-time hires.

So what is OnlyCFO’s compensation?

I obviously should charge a high premium for my time :), but like I said…you can’t just assume 1% ownership (90th percentile per Carta benchmarks) because I said it’s a pre-seed company.

Why? Two main reasons:

• How hot is the company? Not every pre-seed company is the same. Some have zero traction and unproven founders while others have real traction and a proven team. We have seen examples of this with some AI companies that go and raise $100M+ pre-seed rounds at $1B+ valuations…

• What are the hours and scope? Unlike full-time employees where there is clear scope/expectations (40+ hours per week), the scope and time commitment can vary wildly for advisors. This makes equity benchmarks less insightful for advisors compared to full-time employees.

Advisor Agreements & Scoping

1. What Do Advisors Do?

There are a few reasons a company may hire an advisor. Some reasons are better than others…Below are the main ones:

• GTM advisors - People who have a large distribution in a company’s target audience, deep connections at important prospects, or someone who is an experienced GTM leader for your market/ICP.

• Technical advisors: Help guide, shape, and recruit for the product. These folks often have good connections at partners or prospects.

• Industry experts - Deep industry expertise to help guide the company

• Network - Help recruit, intros to prospects, or intros to investors

• Credibility boost: Looks cool to have a big, credible name as an advisor. Founders won’t admit that they do this, but they do. But this might be the worst reason to hire an advisor. They require the most equity and usually do very little. You might be surprised how little VCs actually care about your list of advisors…

2. Set Expectations and Scope

Often the problem is that companies bring on an advisor without any real expectations or scope of what that advisor will do.

• What are you hiring an advisor for?

• What are the expectations?

  • # of calls per month

  • Some kind of regular output

  • Providing regular network introductions

  • 💡Tip - I don’t like setting specific hour requirements or being too descriptive of exactly what the advisors will do. They are not consultants paid by the hour. But you can/should be clear about general expectations from the relationship.

• What does success look like? What are the targets?

Without proper scoping, you won’t really know how much equity they should be given. Someone doing weekly calls or providing important prospect connections should probably get more equity than someone just doing monthly calls (all else being equal).

3. Defining Advisor Equity Terms

Equity compensation for advisors looks different from equity compensation for employees. Advisors get much more favorable terms:

• Vesting Period: Typically vests over 2 years (or shorter). Advisors are most helpful until the next level of scale is reached when more full-time folks are hired to fill gaps so a shorter vesting period than employees typically have (e.g., four years) usually makes sense.

• Vesting Cliff: Unlike employees who almost always have a 1-year cliff, advisors typically have no cliff or it’s very short (like 3 months) because equity is their only form of compensation.

• Early Exercise Provision: Most advisor grants have an early exercise provision since it can be very beneficial to exercise early and the price is usually very small at early company stages (where most advisors are hired).

• Equity Type: Typically stock options (NSOs since they aren’t employees), but RSAs are sometimes done when the company is very early.

• Single Trigger Acceleration: Advisor shares are often accelerated in a change of control event (single trigger). Company execs often get double trigger acceleration clauses - there is both 1) a change of control and 2) they are fired (without cause). Most other employees (non-execs) rarely get either of these acceleration triggers.

• Post-Termination Exercise Period (PTEP): Often this is 90 days just like employees, but with one important caveat. Advisors are rarely actually terminated, even if they stop providing work, so their PTEP never actually starts so they effectively have until the grant expires (10 years).

  • The advisor is fully vested anyway so the company lets them stay active and the PTEP clock never starts

  • CEO forgets about actually terminating them

  • Company doesn’t want to create conflict

  • Founder is embarrassed to terminate an advisor when they discover they aren’t a fit. Most advisors are well-connected so it’s less friction to just stop talking to them. This is why you want to be REALLY sure you want the advisor first.

Advisors are like barnacles. They grow on your company and every once in a while they need to be scraped off. — Dave Kellogg

When I first got the keys to the cap table I was shocked how many “barnacles” we had. Still vesting, but not doing anything. Monthly/weekly calls stopped a long time ago.

Be generous with advisors, but set vesting terms and equity size relative to the time period you think you will need the advisor. Think carefully before hiring any advisors because terminating them won’t be easy.

How do I become an advisor?

Start with these two questions:

1. “Why would a company want you as an advisor and spend their precious equity to compensate you?”

2. “What can you solve as an advisor (working a few hours a month) that the current team can’t figure out by themselves (especially with AI)?”

A deep network is the most common reason I see for hiring advisors (especially today). Product expertise can also be very important when the early team doesn’t have deep product knowledge in some part of the product (can massively improve roadmap and building the right thing).

A startup isn’t typically searching LinkedIn or opening up a recruiting search to find an advisor. Advisors come through the network of the founder, exec team, VCs/board, etc.

Go network…pro tip: being great at what you do makes it easy to build a strong network

Bonus Content: Did a Cursor Advisor Make $384M?

SpaceX finalized its $60B acquisition of Cursor this month, a 4-year-old company.

If you were lucky enough to get an advisor role with Cursor by the pre-seed round, you might be taking home ~$384M for maybe a few hours of work a month for a few years…

Here is my quick math:

After the $60B acquisition was announced, several people posted DMs from Cursor’s CEO, who was asking for advice while he started Cursor as an MIT student…Alex’s $600M missed advisor payday is inflated because it doesn’t take into account fundraising dilution, and tbh, the 1% is probably high. But could he have gotten 0.5% and made ~$180M? Maybe…

There are VERY very few Cursor-level outcomes in the world.

Final Thoughts

If you are going to hire advisors, then make sure you carefully scope it, set expectations, have a good agreement, and provide an appropriate amount of equity based on all of this.

I look forward to becoming filthy rich from my advisor shares, but in the meantime I am excited to just have some fun and learn with great founders. I will be sharing some of the journey from this company in future posts (and will eventually name them).

Subscribe now

Advisors (and employees) dreaming about Cursor-level outcomes…

Footnotes:

• Get a Free Sales Tax Analysis - talk to my friends (and sponsor) Numeral so you don’t screw up fundraising or an acquisition because of sales tax.

• Here is the FAST agreement for advisors. A lot of companies use this agreement and they just released version 3 last month that has some good updates.

*Disclosure: Nothing discussed in this post is financial, tax, or investment advice.

CFOs are pushing AI adoption in finance but are their teams, processes, and systems actually ready to make it work?

Hear what 250 CFOs said about adopting AI in their departments in Rillion’s 2026 Finance AI Report webinar. See what’s working, where CFOs are finding the best AI use cases, and what still isn’t ready for prime time.

2026 Finance AI Report Webinar

I asked Nate Nasralla, CEO of Fluint, to guest write about how AI has changed (or should be changing) GTM efficiency and how we evaluate sales reps. Nate has spent 20 years in sales and currently spends all of his time helping sales teams adopt and leverage AI. So he knows what he’s talking about…

His concept of “floating highways” is my new favorite analogy for how we should think about real AI adoption.

So, on to Nate’s post…

Floating Highways: How to Rethink GTM Efficiency & Rep Evaluation for the AI Era

I’ve been at 18 customer onsites since the start of the year, and there’s a simple analogy I’ve used at almost all of them to describe what AI x GTM is right now.

Remember The Flintstones? How Fred traveled around?

A pedal car, powered by his feet. Pure leg day, every day baby.

Now, remember The Jetsons? They had hovercars, a fundamentally different technology. But what’d they do with it? They “drove” down floating highways. With stoplights, cross traffic, and lanes. Taking tech that unlocked entirely new planes of travel, and bolted it onto what they already knew: roads.

Most GTM teams today are running on floating highways. Somewhere in-between the early 2020 motion, and, what’s to come.

You can run all the “AI Maturity Assessments” you’d like, but you really only need one question to know whether your org has been building a truly AI-native GTM:

If you killed all your AI contracts and ripped AI out of your GTM tomorrow, what would stop working?

Not “go slower,” or “be less convenient.”

What would break?

To the extent things break and just stop working, you’ve built an AI-native GTM. To the extent it keeps humming along, just slower: you’ve got a floating highway.

Now here’s why I think this question matters: the efficiency metrics you manage are built for one of these two worlds. So if your gauges are measuring the machine that no longer exists, or your GTM team is actively tearing down, then, you’re about to make headcount, comp, and budget decisions based on those readings.

So this write-up has three parts to help frame the shift you’re in the middle of today:

1. Part 1 is an audit. Which gauges on your dashboard were built for the old machine, vs. where GTM efficiency is actually found now?

2. Part 2 is about building the system. Breaking down “compound” metrics into the drivers you can influence, while modeling your GTM in a way that wasn’t possible before. (And if you do this, ripping AI out would break something, like a kill-switch.)

3. Part 3 is about people decisions. Headcount, comp, ramp, eval, etc. How to figure out if your reps are performing, and seeing the RevOps team building the systems behind them as someone who’s equally on the hook for a “quota” of sorts. Which flows downstream from the first two points.

(Part 1) What’s Changing?

Are you tracking MPG on an electric car?

I still drive my 2015 Audi Q5 because it’s a diesel, and gets 35 miles per gallon (better than my wife’s 21 MPG BMW). But, if I go to an Audi dealer and compare it to the e-tron, my Q5’s MPG is irrelevant. The e-tron’s an electric car, it’s an entirely different category.

Similarly, here’s a shortlist of metrics I’d be cautious of using in the coming year:

• Pipeline Coverage: this used to be a proxy for “more pipeline = more revenue.” But generally, mid-funnel conversion rates are the same or declining year-over-year, with more of a buyer’s evaluation pushing further down the funnel, so early-stage pipeline coverage doesn’t carry the same weight.

• Cost per Lead: I don’t know another operator who isn’t shifting their packaging and pricing. We’ve launched two new products in H1 at Fluint, each with a slightly different pricing model and GTM experience (PLG, self-serve with consumption bands for technical RevOps users, vs. core sales-led experience for non-technical buyers).
  
  So my old measures of who a lead is, and my cost to “buy” one more, don’t work for AI-native customers coming in self-serve.

• Activity per Rep: this used to translate to pipeline. But now, activity is cheap and essentially infinite. So channels are saturated, and creating truly qualified opportunities is about a smaller set of targeted activities, born from a rep’s human judgment.

The common theme here is that sales efficiency is now found deeper in the mid-funnel:

AI is compressing the seller’s work (prep, follow-up, CRM fill, etc.), but, it largely hasn’t touched the buyer’s internal process after top-funnel research. A buyer still has to manage all the internal comms, stakeholder drama, and nuances of human approvals, to either expand a bottom-up experiment into an annual commit, or, for a customer to move to a new price and package.

As an example of this, let’s take a company that’s:

• $100M ARR with three core segments.

• Enterprise ($45M ARR, ~$250K ACV, with 20 AEs).

• Mid-Market ($35M ARR, ~$60K ACV, 25 AEs).

• SMB ($20M ARR, ~$12K ACV, 12 AEs).

We’ll call them “Bedrock” software. (My creation with illustrative numbers, but a composite of a few of our customers.) Their FY24-26 shows the evolution of (1) pre-AI, (2) the floating highway transition, and then, (3) refactoring for the AI era.

FY25 looks pretty good here when you focus on the top-funnel metrics, like lead generation and pipeline coverage. But those top-funnel gains don’t translate to a 1:1 revenue efficiency, if you follow the quality of the pipeline down through the middle and bottom-funnel, like Stage 3 conversion rates and sales cycle times.

Which is what their FY26 is correcting, with less focus on topline effort metrics (when effort was scarce and roughly equal to results). And more focus on mid-funnel efficiency, like stage-by-stage conversion.

Revenue comes from a mix of humans and agents, but agents don’t carry a quota.

In the past, you could use a simple Quota-Attained : OTE ratio to say, “2-3X is bad, 5-6X is good.” (OTE = on-target earnings, what you pay a rep in variable and base compensation).

This compares the human cost to carry out a set of sales actions, to the revenue they produced. But now, a growing share of the sales activity is agentic vs. humans alone:

Which means every new dollar is increasingly system-enabled, not just sales-enabled. And the revenue generated by a system vs. sales carry different cost profiles. Now, follow me here:

• You can split out the cost of each pretty quickly (token spend vs. payroll).

• But it’s not so easy to attribute a percentage of a revenue outcome to an agent vs. a human rep. (The first half of the “Quota:OTE” ratio.)

• So, if you’re spending on tokens to drive up the rep’s efficiency, and to take on a greater share of their work, shouldn’t their quota go up? But by how much?

• And if you’re spending legit dollars on those agents’ tokens, don’t you need to know what the agents return? How well do they “attain” their “quota”?

The point we’re getting to here isn’t about capping AI spend (”tokenmaxxing” critiques have lost the plot). It’s more about the same question that marketing raised a decade ago. Attribution:

Which AI dollars trace to specific closed-won revenue, and which are just… waste?

This same concept around attribution needs to migrate further down the mid-funnel.

Where the next lift in efficiency is migrating (see point #1 above), and agent contribution and spend are growing the fastest. And once that “instrumentation” is in place, it’ll change the assumptions you use to drive capacity planning, compensation, and 2027+ budgets.

Which brings us to part 2, building the system, before we get to part 3, people decisions.

(Part 2) Building the System

Model your GTM:

Your FP&A team models the business with a “driver tree.” But GTM never got the same driver-based approach, because deals seem “non-deterministic.” Too many variables, and too much gut feel from sales leaders who need to sit on weekly forecast calls to divine whether Lakers season tickets are enough to swing Deal A our way.

But now, you can use a lot of data to model the unique properties of your GTM: specific inputs (“behaviors”) that predict that deal, instead of relying on a general rule, “Stage 4 = 80% value.”

“You mean like having our team set up some Gong trackers or something?”

No, I mean a true model that can (1) weight dozens of sales behaviors, (2) retrains itself as your motion changes over time. Which is what you already do in FP&A, but for revenue.

Which you can do for revenue, in two ways:

1. Step one is what you feed the model: it’s about logging all the GTM activity data you have access to, joined to your revenue data.

2. Step two is what the model feeds you: thresholds that signal risk in the pipeline and base. Signals like, “Hey, deals in [ segment ] evaluating [ product X ] close 59% less often when [ title ] isn’t involved by [ stage 3 ].” So Finance doesn’t “peanut butter spread” the same probability across every deal in the same CRM stage.

(This, by the way, is what I spend all day doing with customers at Fluint.)

Break out each segment independently.

I’m not telling you anything you don’t already know or do here. But before we get to part 3, we have to call out that AI delivers efficiency across Enterprise, Mid-market, and SMB segments differently. And we can’t tell an “AI efficiency story” in broad strokes.

For example, at a board-level, let’s say Bedrock’s LTV:CAC ratios looked good after the last year of AI rollouts. CAC drops from $47K to $36K in FY25, so we’d celebrate, “AI efficiency!”

But if we break it out:

The mix just shifted toward SMB as a greater percentage of the customer count, where CAC is structurally lower.

Now, that’s still a pretty high-level example. But the idea is the more granular we go (cost per qualified opp, stage conversions, cycle time, ACV, etc.), the more we’re going to find different takeaways for different segments.

(Part 3) People Decisions (Both Reps and RevOps)

Model the distribution of your reps: how wide is the bell curve, and can you skew it even further?

As a simple thought experiment, say you have only two sellers on your team, and a CFO reaches out to you as a warm referral. Which rep do you intro them to?

• Rep #1: closes 55% of the accounts they work, at $98K ACV

• Rep #2: closes 15% of the accounts they work, at $58K ACV

It’s kind of a no-brainer. Rep #1.

I mean, why wouldn’t you send 100% of referrals their way? Well, eventually, they’ll hit a “productive capacity.” They’ll get tapped out, so adding more accounts to manage drops their close rate and creates diminishing returns.

So instead of pushing your GTM team to “raise the floor” and get underperformers up to quota, first ask: Can we skew the bell curve further to the right? It’s the highest-probability bet.

Similarly, ask: What changes did we make to our GTM process, to drive up our reps’ “productive capacity?” How many more accounts are they working?

I’d much rather lean into the contribution we earn from more variable pay to high-performing reps who’ve cleared their base + benefits, vs. grow off a larger number of average reps.

Ramping new reps is now an infrastructure vs. a talent question.

The old way of ramping a sales hire was a big PDF playbook, updated once annually at SKO.

The new model is a “living” system that rewrites itself based on the GTM data that’s changing every week. More “flywheel” than “process.” Which just couldn’t have existed 18 months ago.

The idea here:

Your top-performing reps at the front of the bell curve (above) often “work in the dark.” They have a gut-feel that drives them to make different decisions, and that judgment isn’t captured anywhere. So when they leave, that knowledge leaves too.

But that judgment is as valuable as any asset capitalized on the balance sheet today.

If this sounds like an operations problem, in my experience, it’s a comp problem, too.

• With AI, now, you can build a model that makes those subjective judgment calls from top performers an explicit dataset.

• That model is constantly trained on your actual GTM motion, not the market average (a general LLM), to guide new hires.

• But first, you need to trace a top rep’s set of behaviors, with them actively contributing training data back into the model.

• To make sure they do that, create a SPIF that gives top reps a payout based on their data added to the model. Which is them helping you build an asset, that’ll appreciate over time.

For example:

Let’s say a high-performer comes off an exec-level customer meeting, and the next step is to loop in the customer’s procurement team. Well:

• Deal #1: the procurement intro comes in at 8:30pm same day, Friday night

• Deal #2: the intro lands the following Wednesday, business hours, at 10am

Those are WILDLY different signals on the level of priority. Which will change how a top rep approaches pricing and negotiation, to hang onto deal value, maximize commitment terms, etc.

The model traces that change in behavior, and feeds direction to a new hire to act similarly. Which moves them closer to the ACV, win rate, etc. of a top hire.

Ultimately, that means you’re funding and grading RevOps and Enablement based on their ability to compound revenue in a system, which should show up in your rep cohort data:

Did the last class of sellers ramp faster and hit a higher level of revenue-per-rep than the cohort before them?

Capacity planning isn’t “linear math” any more.

It’s more like a multivariate equation.

You used to say, well, “We need X ramped reps, with Y quota, to create Z sales capacity…” But based on the above, there’s a whole set of variables changing beneath that.

Like how far the bell curve is skewing, how much the productive capacity per rep is changing, how quickly ramp time is compressing, how far the span of a leader’s control is expanding, etc.

My old board chair, Lee, used to tell me, “Sometimes Nate, you’ve got to sustain the pain to find the gain.” And what he meant was that we deliberately had to understaff to see how far we could push the system.

I’d suggest FY27 planning is a ripe moment for this. With a choice as you add sales capacity:

• What % of sales capacity do we expect to add from new headcount?

• What % do we expect to add from “system leverage”?

• How does that split compare to last year?

I’d expect that split to skew more towards the system than the seller, toward revops and not just reps, as more dollars are used to expand top rep capacity, vs. just seats. (And as I’d come to find out, the regret cycle of an unnecessary hire is longer, and comes with severance…)

What did you change in your rep’s job description? In RevOps’?

Sales Reps:

As an example of what I mean, earlier this year, after we’d rolled out comp plans, we realized that a more technical product mix (which starts self-serve before a sales-led motion), doesn’t lend itself to the typical, “AE sells the logo,” then “CS onboards and retains.”

Instead, we needed to move to a single account-owner structure, with technical enablement. Where Account Directors own the account from first contact through full customer lifecycle.

Which meant cutting our CS org (hard but necessary), and now, judging our ADs (and paying them) based on a combo of new logo activations, upsells, and retention.

Which forced a few “back to basics” questions for us:

• Does the comp plan match the job description?

• Where it doesn’t, what sales behavior do we want to measure?

• And, what “kind of dollar” is worth more? First spend from a new logo? Dollars committed in a multi-year agreement? What will it cost us to get that specific dollar vs. any dollar?

Here, it’s worth noting reps generally optimize for whichever dollar is the path of least resistance. So if the price you pay for each dollar is equal, they’ll figure out a form of “commission arbitrage,” moving toward the segment that’s easiest to get new commission dollars, to exchange them for what feels like a higher payout.

Which you can rebalance with a type of dollar “taxonomy” (net-new / expansion / renewal, by segment) based on the CAC / margin / strategic value to you.

RevOps:

Back to one of the earlier questions we asked: if AI agents are expected to enable an increasing share of revenue, but they don’t carry a quota, how do you measure them?

Or more accurately here, how do you comp their creators?

The RevOps job description is splitting into two directions right now:

• AI: The engineering function responsible for architecting systems that both learn and execute (creating new data to learn from in a loop).

• BI: The reporting and rules functions to dashboard the outcomes, and ensure there’s consistency in operating rhythms across the business.

As an example, Scott Gibbs from Meltwater describes the new RevOps functions they created to enable this. The revenue engineers need upside, and while the comp norms are still being established, the one question I’d have for you is:

• Is the RevOps org on the hook for driving a greater share of the variable revenue outcome?

• If so, to what extent will they share in the upside of a good outcome, with variable comp attached to it? More than last year?

To be clear, I don’t think this looks like a commission plan for RevOps. But it does look like a direct conversation on how their role and compensation are being reorganized.

Slope, not intercept.

The rising tide of AI does not float all boats. So while it’s good to evaluate the efficiency gain AI brings to the revenue org as a whole, I’d suggest a more granular, rep-level look.

It helps to evaluate each rep as a “line, not a dot.”

Buying behavior is changing fast, and a seller’s ability to evolve with it isn’t guaranteed. (In case that wasn’t already obvious.) So, my suggestion is to create a simple line chart of trailing Revenue:OTE on one axis (the “intercept”), with their trajectory over recent quarters on the other (the “slope”).

Then, you can split that grid into quadrants:

• Who were previously the bottom-half performers, that have been accelerating fastest? (Likely a future asset that an annual stack rank may have thought to begin managing out, but, could be among your rising stars.)

• Who were previously the top-half performers, whose performance may be “decaying” over time?

• And collectively, is the median trend line up, or down, midyear? Is the “system” working?

So, where do you go now?

Back to Fred and his pedal car:

The teams still running on floating highways will treat AI as a line item to cap, and they’ll get what they measure: the same motion, just a little faster, with a new expense to explain to the board.

Vs. the teams refactoring for the AI era see it differently. Instead of bolting new tech onto the old roads, they’re building a system where ripping AI out of it would mean breaking the model.

So if you’re somewhere in-between the early 2020 motion and passing the “kill-switch test,” here’s where I’d start this week:

• Check your dash. Was this gauge built for the old machine? Anything measuring topline effort (when effort was scarce and roughly equal to results) goes on the watch list. Move your attention to mid-funnel health, like stage-by-stage conversion.

• Make FY27 planning about the “multivariate” equation. As you add sales capacity, force the question: what % do we expect from new headcount, what % from system leverage, how does that split compare to last year, and what operational changes did we make to drive it?

• Evaluate the revops job description, and sales reps as “lines, not dots.” Plot trailing Revenue:Quota with a quarterly trajectory, find the rising stars, and over-invest in the top quartile of reps. While confirming your revops team is comped and focused on building the system beneath them.

Footnotes:

• Join me and attend Rillion’s 2026 Finance AI Report webinar to learn how other CFOs are implementing AI

• Subscribe and share the OnlyCFO newsletter with your teams

Subscribe now

*Nothing in this article constitutes legal, tax, or investment advice. I am not a lawyer or tax accountant.

Register for Rillion’s 2026 Finance AI Report webinar to hear what 250 CFOs are saying about implementing AI in the “Office of the CFO”.

What’s actually working, where the gaps are, and where things break.

Every finance team is trying to become more AI-native. Hear from AI-forward CFOs who have done it — and learn what worked, what didn’t, and what they’d do differently.

2026 Finance AI Report Webinar

The Most Common Cap Table Issues

A few years ago, I was working with a company that had the worst cap table issues I had ever seen. It was bad… If I was the CFO while these issues happened, I would have changed my name and moved out of the country.

The one positive from this experience, though, is that I learned A LOT about these cap table issues (thanks to some very expensive lawyers), how and what issues can be fixed, and how you might remediate them with your employees.

If you have any ownership over the cap table (CFO, accounting, legal, HR, etc), then please read this post and/or share it with your teams that do. It’s also helpful for regular employees to understand how this stuff works too in case it happens to you.

Mistake #1: Unapproved Grants

An admin error causes grants to not get included on the list that the board of directors approves:

• HR fails to tell finance/legal (or whoever owns the grant list) about promotion grants

• There is a timing cutoff error on new hire grants so a group of new hire grants don’t make it on any approval list

• A founder promises equity but forgets to formalize and communicate with HR.

• Someone forgets to actually add it to the board meeting agenda so the batch of grants never gets approved (everyone just thinks it got approved)

What’s Wrong:

No board approval = no grant.

Offer letters always say something to the effect of “contingent on board approval” with respect to the equity listed in offer letters. This is because typically only the board of directors can approve equity grants.

I have even seen cases where the equity admin thinks the board approved the grant since the board meeting happened, stock option agreements are created, and then sent out and signed by employees….but the board never approved the grant so it still isn’t a legitimate grant. And the employee would have no way of knowing that the company screwed up on the paperwork.

The Fix:

For most equity-related issues, waiting to fix it will only compound the problem.

• How much higher is the 409A fair market value (FMV)? Companies can only grant at (or above) the current FMV so if your company has gotten more valuable, granting the same number of stock options means the employee will be worse off since there is a higher strike price to buy the stock options. This is less of an issue with RSUs.

• Have they exercised an invalid grant? If the issue has gone unfixed for long enough, then some invalid grants may have already been exercised…This makes it much more painful. This is harder to unwind because you have to fix the cap table, refund exercise prices and figure out what to do with any taxes that were paid (AMT, payroll, etc)

• Are they still employees? Now this is where it gets really painful. “Hey John, remember how you exercised $50K worth of stock options last year? Well, funny story….those shares you think you own aren’t real.” Former employees are never as kind to you as the current ones. Also, once someone leaves the company, there are a lot fewer options to remediate the problem.

There are a few imperfect fixes for unapproved grants:

• Re-grant at current FMV. But like I said, if the valuation increased, the employee gets shortchanged. If the 409A decreased then I guess it works out. Slightly bittersweet that the company valuation is decreasing though.

• Top-up grants. Give more options to make the math work. The employee might be happy, but more stock options at a higher price is only good if they expect the price to increase even more than the spread they lost between what the FMV price should have been at versus what they got it at. The company/board is unhappy though because this means more dilution.

• Restricted Stock Awards (RSAs). Can help patch past mistakes, but come with tax consequences that can also create problems for employees that may also need to be solved.

• Cash bonus / settlement: Give some cash to make up for potential tax consequences and/or change in FMV. This is the main fix if they are now an ex-employee that exercised their “options” that don’t actually exist.

• Advisory role workaround for ex-employees. If the employee already left, you can hire them as a contractor/advisor so that you can issue equity to them as a “service provider.” There is some risk here since they should actually be providing some kind of advisory services, which is why this is usually reserved for large grants given to VPs or above.

• Return exercise price. If an employee exercised an invalid grant, then you have to return the exercise price and you also have to decide what you will do about any taxes they paid (AMT or payroll taxes on NSOs). You might also be able to do a W-2c if it was within the past three years, but that is always painful so a settlement payout is often easier.

The only potential “get out of jail free” card is if the invalid grant was issued to an employee that has left and never exercised their stock options. For those, you may have dodged a bullet…

Mistake #2: The Stale 409A Valuation

Private companies must get a new 409A valuation at least every 12 months or following any material business changes (most commonly a financing round).

• Forgot it’s been 12 months since the last 409A was effective and granted equity.

• Financing term sheet is received, but deal isn’t closed. The date the term sheet is received likely invalidates the prior 409A

💡Common mistake: 409As are valid 12 months from the 409A report effective date. Not the 409A report date…. The 409A might be effective on 5/1/25 but your 409A advisor finishes the report on 7/1/25. It expires on 5/1/26 (12 months from effective date)

What’s Wrong:

If a stock option strike price is granted below FMV (regardless of whether its an ISO or NSO), it’s subject to Section 409A, which is a bad thing... There are lots of taxes and penalties for employees if this happens, including losing the ISO status if it was granted as one. It is NOT something you would ever want to do intentionally. It’s a nightmare for both the employee and employer.

The Fix:

Stale 409As are different from unapproved grants. Unapproved grants never actually existed. These grants might exist but they are wrong (and create tax issues). Here is what the fix may look like.

1. Cancel or deem invalid:

   Cancel and “deem invalid” is kind of the same thing, but deeming it invalid is cleaner I think. Basically all stock option plans say that grants must be done at (or above) FMV. So a company can just say the grants didn’t follow the plan so they weren’t properly granted and therefore don’t actually exist. If you “cancel” them then you are kind of saying they were valid, which may mean there should have been tax consequences for below FMV grants.

2. Get a new 409A:

   You probably need to update your 409A. While you may deem the grants invalid under the stale 409A, you still may have a problem with valid exercises that happened under a stale 409A…Exercises have tax consequences depending on the FMV at the time of exercise so if the 409A was stale at the time of exercise then that is a problem for them as well which may require adjustments for tax purposes.

3. Make employees whole:

   Since the grant doesn’t exist the company now has to do one of the make-whole options that I mentioned previously (re-grant at current price, add more options to make up for the increase in FMV, cash bonus/settlement, etc)

Don’t forget to refresh your 409A. The two key times are: 1) at least every 12 months and 2) after a financing. But also understand at what point the 409A goes stale and you must stop any new grant approvals (receiving a financing term sheet, getting an acquisition LOI, etc)

Mistake #3: The Zombie Exercise

This happens at lots of companies… Most stock option agreements have a 3-month post-termination exercise period (“PTEP”). This means that once an employee terminates they have 3 months to exercise their option. Otherwise, the stock option expires.

Here is what usually happens:

1. Someone forgets to enter the termination date into the equity management tool

2. Three-month window never starts in the system

3. Employee checks 6 months later and decides to exercise because it’s still available in the system

💡PTEP vs ISO Status: A lot of people incorrectly believe that stock option PTEPs must be three months or less for stock options to get the favorable ISO status. That’s not true. You can grant stock options with a 6-year PTEP and still have ISO status.

The rule is that stock options just have to be exercised within three months to get ISO status. If someone exercises after 100 days post-termination then the stock options just automatically flip to NSOs.

Having worked at a company that granted long PTEPs, this can be an awesome benefit that employees love, but…it 1) adds a lot of administrative burden and 2) it can add a lot of dilution overhang which obviously means everyone’s slice of the pie shrinks (~50%+ of people don’t exercise stock options when they leave, but if you have a long PTEP then they don’t have to).

What’s Wrong:

If the plan says the option is expired, then it can’t be exercised (regardless of what the system says). The employee thought they owned shares when they exercised, but they legally own nothing.

The company just owes the employee their money back…

The issue may have been caused by a company admin mistake, but the stock option plan is clear, which the employee signed, so the company isn’t typically obligated to do anything (but perhaps legally questionable if the exercise was accepted by the company).

The Fix:

This is different than the first two mistakes because it is kind of the employee’s fault no matter what. They signed an agreement that they only had so much time to exercise options after terminating. An admin mistake doesn’t change that.

Most frequently I just see the company refund their exercise price and undo the “shares” in the system.

But some companies feel bad and try to do something (cash settlement, advisor agreement to regrant shares, etc).

Mistake #4: Waiting to Fix Cap Table Issues

For some issues, it doesn’t matter if you wait to clean them up. Cap table-related issues are not one of those…The longer you wait, the worse (and more expensive) they usually get.

And cap table issues always surface at the most inconvenient moments:

• During M&A diligence because potential acquirers are poking holes in everything.

• During financing diligence when VCs are making sure they understand who owns what.

• During IPO prep while legal is validating ownership before going public

• When an employee asks some questions and someone actually digs into the details.

Don’t wait. Make sure your cap table is clean now. If it isn’t, spend the time cleaning up the issues now. It will only get more expensive, the longer you wait. I have seen some big purchase price adjustments in acquisitions because of issues discovered here…

The Cap Table Checklist

Here are the additional checklist items you need to implement so your cap table doesn’t delay/kill a fundraising round, an acquisition, or an IPO. It will also save you lots of money on legal cleanup fees and embarrassment with your board...

1. Board Approvals or Bust: No approval, no grant. Make sure you have a good process for double-checking the grants that should get included (HR, finance, legal, and CEO final review). I use AI now for this…”Claude, compare approved grants from board meeting to pending grant list to new grants in Carta.”

2. Fresh Valuations: Set calendar reminders on month 10 or 11 to email your 409A advisors to kick off the process. And make sure that your finance leader who is in the know about fundraising and M&A knows the rules around 409A valuation timing.

3. Expiration Tracking: Two things I do for this, 1) get a good admin to set a regular cadence for updating the equity admin tool for terms and 2) monthly/quarterly reconcile all terms per HRIS with all terms in the equity tool.

4. Good Equity Admin Tool – this will make your life a lot easier.

5. Monthly/Quarterly Reconciliation – Reconcile the cap table with board minutes and HR records. “We have a new hire from 6 months ago that hasn’t received a grant yet”

6. 83(b) Tracking for Early Exercises / Restricted Stock: Something that always comes up in diligence is if proper 83(b) tracking has been performed. Yes, it’s the employee’s responsibility, but companies should keep a record of the filing too in order to avoid issues with the employee and potential liabilities.

7. Track Early Exercise Terminations: I have also seen companies just forget about folks who early exercised options and then leave before their full vesting period. If the company doesn’t repurchase those shares then they may forfeit their right to repurchase them and then the employee gets to keep the shares even though they didn’t vest.

8. Track Stock Option Pool: I show the remaining pool and 12-month grant forecast to the board every time we get new grants approved. It’s easy to add and keeps everyone aligned on dilution and the remaining pool available.

9. ISO vs NSO Classification: Make sure your admin and equity tool handles ISO/NSO rules correctly. For example, contractors/advisors can’t receive ISOs, large grants can run into ISO limits, and post-termination exercise periods can change the tax treatment. Get this wrong and the company will report and pay the wrong amount of taxes.

10. Rule 701 Tracking: If you don’t track this and apply exemptions when appropriate (like Reg D or 4(a)(2)) then you can accidentally cause a ton of work and broader disclosures of financials…Make sure your legal firm (or someone) is tracking this.

Final Thoughts

A poorly run cap table will cost you lots of money in legal fees, embarrassment with the CEO and board, piss off a lot of employees, and waste a lot of time for many people.

Lawyers can paper some mistakes away (“Section 204” has saved me from a few embarrassments), but a lot of equity mistakes are not easily fixed and cost a lot of money.

Make sure your cap table is clean now before you are in diligence of some kind and it pops up at the worst possible time. These things take time to fix…

Footnotes:

• Join me and attend Rillion’s 2026 Finance AI Report webinar to learn how other CFOs are implementing AI

• Subscribe and share the OnlyCFO newsletter with your teams

Subscribe now

*Nothing in this article constitutes legal, tax, or investment advice. I am not a lawyer or tax accountant. The above article is based on my personal experience and actions that I have seen performed. Consult with your legal team on what to do based on your circumstances…

Sales tax is coming to California (January 1, 2027)! This is bad news for all of us tech folks…

• A high tax rate between 7.25 and 10%+ (depending on jurisdiction)

• Your CA-based customers now get hit with sales tax

• CA-based companies have a big increase to all their software spend

More than 3,000 businesses use Numeral to get sales tax right. Mess up sales tax and you will end up with a huge liability. Get a free sales tax analysis 👇

Get Free Tax Analysis

SpaceX Stock Price Tanks

SpaceX IPO’d last month at a $1.7 trillion valuation ($135/share) and briefly touched $3 trillion ($226/share) a few days later on June 16th, which was the same day they announced their $60 billion acquisition of Cursor.

SpaceX stock has been falling ever since…down 50% in the past 45 days.

Guess who is cheering as SpaceX stock price crashes?

Cursor investors and employees…

Wait…but Cursor is being acquired by them. Why would they be happy?

Because the deal hasn’t officially closed yet (expected this quarter) and based on the deal terms, Cursor is benefiting as the stock price falls.

Why SpaceX’s $60B Acquisition of Cursor Is Getting A Lot More Expensive

April 2026: SpaceX’s Call Option to Buy Cursor:

A lot of the diligence and pricing was done prior to the acquisition being announced in June because SpaceX signed a call option agreement for the exclusive right to buy Cursor for $60B if they decide to move forward.

A call option agreement like this is pretty unusual. The option gave SpaceX time to complete its IPO and evaluate the acquisition before making its final decision.

The option wasn’t free for SpaceX. There was essentially a $10B breakup fee if SpaceX called it off…

June 16, 2026: Exercising the Call Option:

On June 16th, SpaceX announced the exercise of the option and entered into a definitive merger agreement with Cursor. At this point, a lot of the acquisition work was already done as part of the call option agreement.

But the deal doesn’t “close” when you enter into a definitive merger agreement. SpaceX said that they expect the deal to close in Q3’26 (so in the next couple of months). The primary reason for the gap between definitive merger agreement and close is the antitrust stuff (like HSR).

Buyer vs Seller Risk

The acquisition price wasn’t $60B in a suitcase full of cash. It was $60 billion of SpaceX stock.

But $60B of stock as of when? This point matters a lot…

There are two primary ways stock deals can be structured:

1. Fixed Exchange Ratio - The number of buyer shares is fixed, so the dollar value rises or falls with the buyer’s stock price. (Risk is on the seller)

2. Floating Exchange Ratio - The dollar value is fixed, so the number of buyer shares rises or falls with the buyer’s stock price. (Risk is on the buyer)

Fixed is by far the most common deal structure. In fact, floating exchange ratios make the most sense when the buyer is public with an actively traded stock.

Also, when floating exchange ratios are used, they rarely come without a “collar.” A collar limits how much the exchange ratio can adjust when the buyer’s stock price moves up or down. Without a collar, if the buyer’s stock price crashes, the buyer could have to create an unlimited amount of dilution to settle the acquisition.

SpaceX agreed to buy Cursor for $60B using an uncollared floating exchange ratio based on the volume-weighted average closing stock price of the seven days prior to the deal closing.

Remember, the deal was announced in June, but it isn’t expected to close until sometime this quarter (hopefully). While Cursor knows that they will be getting $60B worth of SpaceX stock at closing, SpaceX still doesn’t know how much dilution the acquisition will create for them…

Date TBD: Settling $60 Billion Acquisition Price

I actually got the transcripts from SpaceX’s recent board meeting…

• Elon: Buying Cursor for $60B is a good bet. It will accelerate our AI vision.

• Board: It’s a lot of money, but I guess a small % of your market cap

• Elon: Exactly! This $80B acquisition will be the most important $100B we have ever spent.

• Board: Wait…I thought it was $60B?

• Elon: Look, if we don’t spend $125B buying Cursor then we will lose in AI

*Obviously the above is fictional, but it’s what I imagine everyone is thinking during SpaceX’s collapsing stock price.

SpaceX’s stock price has fallen by ~52% since the deal was announced at SpaceX’s peak valuation. So the floating exchange ratio for the $60B of stock is now 2x more expensive from a dilution standpoint since the deal was announced. And probably feels like they are now paying ~$125B for Cursor…

The acquisition still hasn’t closed yet so maybe the stock could recover a bit but it could also keep falling…

Here is the potential vicious cycle that can happen:

1. SpaceX stock price drops (for whatever reason)

2. SpaceX effectively pays more to acquire Cursor (more dilution)

3. Stock market drops (has been happening right now) so investors believe Cursor might be worth less than $60B

4. SpaceX stock price drops further because it is locked into overpaying for Cursor (or there is a huge breakup fee) so the valuation is dinged again

5. SpaceX pays even more to acquire Cursor as stock price falls

6. Repeat…

A floating exchange ratio creates buyer risk that could spin out of control.

Elon Hates It; Cursor Loves It

Elon and the SpaceX team hate that the stock price is falling for obvious reasons…but all else being equal, Cursor equity holders are silently cheering.

Why?

If nothing material has changed at SpaceX, then it is still the same company with grand ambitions to be “more valuable than Earth”. But Cursor would rather receive more shares at today’s much lower valuation. If the deal were priced today, Cursor would receive roughly twice as many shares as it would have at SpaceX’s June 16 peak.

Cursor folks are just praying that it doesn’t recover before the acquisition closes…

Fixed cash/exchange ratios are the most common, but as you can see below, if your stock price increases, then the fixed ratio works against you. If Google acquired Wiz in a similar structure as SpaceX<>Cursor, then they would have effectively paid $16B less because Google’s stock price rose 93% between announcement and the deal closing (so they would have had to issue almost half the number of shares than at the announcement date).

Buy Low. Sell High

Companies want to buy low and sell high with their own stock:

• Use equity (e.g. employee equity and acquisitions) when the stock price is overvalued

• Use cash when the stock price is undervalued

This applies when issuing stock-based comp to employees, performing buybacks, and buying companies. But that is hard to time right, especially when you have a floating exchange ratio and a volatile post-IPO stock…

So as long as nothing really changes with SpaceX’s business and its stock price remains deflated through the close, Cursor investors and employees are really happy since they would get a much better price with more room for upside.

Footnotes:

• Get a Free Sales Tax Analysis - Sales tax on software is coming to California so use my friends (and sponsors :) from Numeral to get ready.

• Want to sponsor? shoot an email to onlycfo@onlycfo.io

Subscribe now

*Nothing in this post should be considered tax, legal, or investment advice. Educational purposes only. Consult an investment and/or tax advisor.

Subscribe now

Domo has been one of the worst performing software companies in the public markets for a long time. They have one of the lowest revenue multiples driven by zero revenue growth, breakeven FCF margins, and a long history of burning a lot of money.

Domo has been contemplating a sale for a while given their broken economics, growing debt, and dwindling cash balance. They hired a banker earlier this year to formally find an acquirer. When they tripped their debt covenant last quarter the lender put a clock on them getting acquired. The lenders basically said, “Sign a purchase agreement by July 31st and close by November. Or else…”

Well, this week Domo finally convinced someone to acquire them (at least part of them). Progress Software, a public company, has agreed to pay $400M for Domo’s assets.

Unusual Acquisition

Domo is selling basically all of their assets (software, customers, IP, etc) in a $400M asset sale. Not a typical equity sale.

Why structure it this way?

Domo is essentially creating a SPAC-style holding company that will have ~$246M of cash (after repaying debt and transaction expenses) that they can use to buy companies with.

But also, there are some attractive tax benefits by structuring it this way…

Losing Money Was Actually Domo’s Plan All Along!

One of my all-time favorite tweets comes from investor Cathie Wood (CEO of ARK Invest) where she basically says, “due to our bad stock trading we have lots of losses that can offset capital gains for many years!”

Tax loss harvesting can be part of a legit strategy, but it is a very funny thing to talk about when her funds were down so hard relative to the market.

Maybe losing lots of money was part of Domo’s grand plan?

Because it’s an asset sale, Domo will retain their net operating losses (NOLs). Domo expects to close the deal with ~$900M+ in NOLs that can offset future taxable income (if they ever have any). They currently have ~$1.3B in NOLs, but some of those will be used to offset gains from selling assets in this transaction.

So now Domo has a decent pile of cash (~$246M) and a public company shell to go acquire a company (similar to a SPAC).

The difference with a SPAC is that Domo will have a pretty massive NOL asset. In theory, they can go buy a profitable business and offset profits for a really long time.

What is $900M of NOLs worth?

Hint: Not $900M

That represents the total losses that can offset future profits.

With a 21% federal corporate tax rate, it is at most $189M (assuming a $900M NOL).

Domo also has state-level NOLs. Their state NOLs are a bit higher than federal NOLs with an implied average tax rate of ~5%. If we assume they have $1B of state NOLs after the transaction, then they potentially have up to $50M of tax savings.

There are two potential problems with the theoretical tax asset that Domo’s shell company owns:

1. Certain NOL rules, if triggered, could partially or completely eliminate NOLs (more on this below)

2. This one is important…the company must actually generate profits for it to matter.

NOL Expirations:

There are two types of NOLs that companies must track because when and how they can be used are very different:

• Pre-2018: NOLs generated before 2018 expire after 20 years

• Post-2017: NOLs generated after 2017 never expire, but can only offset 80% of income each year.

NOL Limitations (Section 382):

Section 382 establishes thresholds that can significantly limit NOLs. Domo’s structure is supposed to help preserve them.

• Trigger? One or more 5% shareholders collectively increase their ownership by more than 50 percentage points over their lowest ownership levels during the prior 3 years. (remember these thresholds because it’s important to something else they implemented)

• Limitation? If the trigger is met, then the amount of NOLs that can offset income each year is significantly limited. If another company bought control of Domo’s shell company (assumed value is $246M) then that company could only use $9.3M of its NOLs each year ($246M * 3.77%). In other words, it would take 97 years to use their NOLs. A lot of Domo’s NOLs were built pre-2018 so those will eventually expire and the remaining NOLs can eventually be used but the present value of them wouldn’t be worth nearly as much.

Cooking Up a “Poison Pill”

Domo announced a “Tax Benefits Preservation Plan” at the same time the asset sale was announced.

The purpose? To make sure no one blows up their NOLs and severely limits their usability.

What is it? It’s a type of “Poison Pill” that essentially stops investors from obtaining more than 4.9% of the shell company’s stock. This poison pill is a right given to every shareholder to purchase Domo stock worth $35 for $17.50. Everyone gets that right except the investor crossing 4.9%. So everyone benefits except them. In practice, once this is implemented, no investor would acquire 5% because they would essentially just be transferring money to all other shareholders. The board can grant an exemption to the poison pill, though.

While the poison pill was written up to protect NOLs (valid reason), it has the side benefit of preventing someone from buying it for the cash and/or public company shell.

What’s Next For “Domo”?

Domo is losing its name and will change its ticker on the stock exchange since all Domo branding is now owned by someone else.

Domo stock jumped 25% on the acquisition news…

Wait…the company is just a shell holding cash and valuable NOLs. Why isn’t it trading at least at the cash per share of $4.84?

It is currently 18% below the expected cash per share (not even considering NOL potential).

It’s common for a transaction like this to trade below cash per share:

• Management risk: Stockholders are relying on someone who burned through $1B of cash to build a bloated business that never made money so why should they expect them to be able to buy a business and run it well?

• Actual cash per share is just an estimate and could change by the time of closing because of purchase price adjustments, transaction expenses, tax stuff, etc

• Still public: Domo’s shell will still be public and will have operating costs of a shell company (legal, tax, management fees, etc) and will have expenses to try and find a company to buy.

• Deal risk that it never actually closes. Risk here seems pretty small.

I don’t know what the Domo shell will buy, but it will have to be pretty small…What can you buy with $246M of cash on the balance sheet? The obvious answer is a broken company like Domo…so if you want something good (growing and/or profitable) it will have to be pretty small. Good chance they take on some debt to buy something bigger, but even so, they still won’t be able to buy something really worthy of the public markets right now with the leverage that a lender would give them.

And remember, they will still be operating as a public company with all the costs that go along with it, which is difficult for a very small company to absorb.

The deal won’t close for another few months, but in the meantime Josh James (Domo CEO) and the board will be on the hunt for what to do with the cash. Become the next MicroStrategy and go all-in on Bitcoin? Buy broken AI businesses and scale them (version of Bending Spoons)? Build data centers? Or just burn more cash and drive the stock price further down?

Goodbye Domo….And I look forward to writing about whatever the shell company does next :)

Footnotes:

• Want to sponsor? shoot an email to onlycfo@onlycfo.io

• Subscribe for more fun and share with your friends

Subscribe now

*Nothing in this post should be considered tax, legal, or investment advice. Educational purposes only. Consult an investment and/or tax advisor.

Subscribe now

Last week, Trump Media (public company that owns Truth Social) announced “Truth API”, which will give customers access to the 10 most “influential” Truth Social accounts in milliseconds, which is faster than everyone else will get the posts.

• The Price? reportedly as high as $100,000 per month…

• The Target Customers? Hedge funds, banks and big trading firms that are dealing with billions of dollars

• The Pitch? Simple…Trump posts crazy stuff that moves markets in major ways so if these firms can get the posts faster, their automated trading algorithms can buy/sell faster (which matters a lot). Also, if they don’t buy it then they are going to be behind other traders (which will cost them lots of money).

Trump’s post (aka “Truth”) about pausing tariffs last year would probably have paid back the $1.2M annual price tag by 100x+ in just a couple of hours for many of these trading firms…

At 1:18 pm on April 9, 2025, Trump posted about a tariff pause and the stock market immediately jumped by ~7-9%. The Nasdaq finished the day 12.2% higher and the S&P 500 9.5% higher.

Imagine being a trading firm not paying to get this info slightly sooner and missing out on one of the largest single-day gains in history.

Obviously, this wouldn’t be for normal retail traders because most investors can’t act on the news (buy/sell) fast enough for the timing difference to matter. But the shops that have spent millions of dollars to automate trades based on the speed of information have to buy this…there is almost no choice.

I am not a lawyer, but apparently there is nothing illegal about this… It is not insider trading, and apparently the President is exempt from the conflict of interest laws that would have prohibited other government officials from doing the same thing.

Does it pass my general ethics sniff test? obviously not…

Unprecedented Levels of Grift

Truth Social isn’t the first social media app to sell faster access to posts, but…

• Insane price: $100,000 per month for faster access to the top 10 Truth Social accounts (really everyone just cares about the one that actually moves markets on a regular basis). This is ~100,000%+ higher than what would normally be charged by any other similar service on a per-post basis.

• It’s for posts from the President of the U.S.: Many of Trump’s posts move markets in major ways and he posts them first on Truth Social.

• Trump owns 41% of Truth Social: More frequent (and market-moving) posting by the president on Truth Social can make him a lot richer because more people will be pushed to buy these $100,000 subscriptions. Because of this financial relationship, Trump obviously financially benefits whenever he posts market-moving things…

And even crazier is that this “Truth API” will probably make up the majority of the company’s revenue by next quarter.

Trump Media only generated $871K in Q1, or $290K a month. What?!? lol. It would just take three customers at the $100K price tag for it to become the majority of its revenue.

Remember, this company has a $2.7 billion enterprise value (once nearly $10 billion)…off of basically no revenue. Nearly the entire valuation is tied to the President of the U.S. and this new product is doubling down on that. There is a limited amount of time (2.5 years) to milk Trump’s presidential status and they are shamelessly going for it.

Is “Truth API” ARR (recurring revenue)?

Well, it’s a monthly/annual subscription that is probably expected to renew for at least a couple of years. But unless Trump gets a 3rd presidential term 🤣…basically everyone will churn in about 2.5 years.

Funny enough though…I think a lot of software companies are in a similar position (ie revenue is no longer very durable) but they don’t have as clear of a churn date.

Definitive churn in 2.5 years breaks the objective of “ARR”. ARR was magical because of its durability. The business model looks much different if “ARR” only lasts a couple of years. LTV metrics will look amazing for 2.5 years…until a wall of churn hits.

Rationally, then, investors would only value this incremental revenue potential based on the temporary FCF increase from it for the next couple of years. But…that would assume whoever owns Trump Media is “rational”.

*Some reports say they are offering discounted prices of $60K per month if customers commit to at least 3 years. I wonder why…🤔. Oh yeah…the only reason people will buy the subscription will no longer matter in 2.5 years.

What Counts as Insider Trading?

Insider trading occurs when three things happen:

1. The trader has material, nonpublic information about a security.

2. The trader buys or sells securities while aware of it.

3. The information was obtained or disclosed through a breach of a fiduciary duty, confidentiality obligation, or similar duty of trust.

Paying for faster access to information is not necessarily insider trading. For example, Bloomberg Terminals, which cost a lot of money, provide a speed advantage, but that doesn’t trip insider trading rules. People are just paying for a faster connection.

Ok, What About Regulation FD?

Regulation FD (“fair disclosure”) is 26 years old and basically says companies must simultaneously disclose information to the public. Meant to protect retail investors from companies sharing material information with analysts in advance of everyone else.

But Reg FD doesn’t apply here because:

• The information Trump is posting about is not about Trump Media.
  Reg FD governs disclosures of material nonpublic information about that company. Trump’s posts about tariffs, wars, interest rates or foreign policy may move many securities, etc

• The API accelerates delivery of published posts but does not provide unpublished information.
  The API feed will deliver posts faster than ordinary push notifications. If the post is already publicly posted and the API just transmits it faster, there is no nonpublic information to regulate.

TLDR: the laws did not contemplate this insane situation… or at least assumed lawmakers would never need to make something like this illegal.

Takeaway: Be Honest and Ethical

I don’t know all the facts on this Truth Social API subscription, but it certainly looks very ethically questionable based on the facts we have (even if not illegal) - creates information asymmetry with retail investors and creates a crazy financial incentive for the U.S. president.

My reminder for business leaders: perform the front-page newspaper ethics test:

Would you be comfortable with your decision if SEC lawyers, your board, your mom, etc. knew about it?

Whether it’s questionable ARR definitions, manipulating timing of customer churn to look better in fundraising, not being honest about potential product incidents in due diligence, etc….

Do a front-page newspaper test and make sure your decisions will pass it.

Footnotes:

• Free month-end close checklist tool: Ditch Excel or expensive tools. Check out this close tool I helped design. Join the waitlist!

• Want to sponsor? shoot an email to onlycfo@onlycfo.io

*Nothing in this post should be considered tax, legal, or investment advice. Educational purposes only. Consult an investment and/or tax advisor.

1. Every parent of an eligible child should open up an account (it’s free). Even if a child does not qualify for the free $1,000 government contribution, there are still meaningful benefits for all other children under 18.

2. Every company should likely create a “Trump Account contribution program”

If you are a CFO, then read below and figure out how to start getting it set up. If you are not in charge of benefits, then send this article to your HR team and/or CFO and tell them they need to implement it :)

Share

*Put your politics aside and get over it being called a “Trump Account”. The program has huge tax benefits and can significantly help your kids. Every U.S. citizen and company based in the U.S. needs to understand this.

Who Benefits From Trump Accounts?

These accounts are early childhood savings vehicles to help kids learn about the value of investing and save for retirement.

• Who? Anyone <18 years old with a social security number (after 18 typical traditional IRA restrictions apply).

• Amount? Total of $5,000 per year (including employer contributions) can be contributed to each child’s account.

• Investments? Limited to low-cost mutual funds and ETFs that track broad indexes mostly composed of U.S. companies.

• Withdrawals? Prohibited until the year the child turns 18 years old and then IRA rules apply (10% additional tax on early distributions with limited exemptions like first-time home purchase, education expenses, etc)

• Free $1,000? Qualifying U.S. citizen born between 2025 and 2028 can receive $1,000 from the government (does not count toward the $5,000 per year limit)

• Taxes? All gains, interest, dividends, etc are tax-deferred (only taxed when withdrawn).

Why Employers Need to Offer It

All the above benefits are great, but one of the biggest potential benefits comes from your employer.

There are three things a company can add some of this benefit (all companies should probably at least do #1 and #2):

1. Enable after-tax payroll deduction which just helps employees automatically invest part of their regular payroll into Trump Accounts. This wouldn’t impact the employer $2,500 limit but would count toward the total $5,000 limit. No tax advantages, but potentially adds a tiny bit of convenience for employees if it helps streamline deposits and limit tracking.

2. Enable employees to contribute pre-tax dollars up to $2,500 to a dependent’s Trump Account. This would reduce their taxable income and save taxes.

3. The company can contribute up to $2,500 per employee which is tax-deductible to the employer and tax-free to the employee.

💡 Note: The combined limit of #2 and #3 is $2,500 per employee (not per child). For example, if the company contributes $1,000, the employee can contribute only another $1,500 through pre-tax salary reduction. The employee could then contribute another $2,500 with after-tax dollars (#1) to reach the child’s overall $5,000 annual limit.

Obviously #3 (company contributions) can add a lot of new expenses for a company, but simply adding the plan and allowing employees to contribute pre-tax dollars (#2) saves employees on their marginal federal income-tax rate.

Pre-Tax Contributions Example

Let’s say I have a household income of $150K and I am in the 22% marginal federal income-tax bracket (I am married filing jointly). If my employer offers this benefit then I can reduce my taxable income up to $2,500 per year and give those dollars to my children’s retirement.

Instead of giving the IRS $550/year (22% * $2,500) you can contribute more to your children’s retirement or you can just save that amount in taxes if you plan to contribute $5K either way. Based on compounding returns in the U.S. stock market (assuming a reasonable 8% growth per year), those tax savings will grow to nearly $0.6M (or about $100K in today’s dollars if inflation adjusted) by the time the child is 60.

So instead of giving the IRS $9,900 over the first 18 years of your kid’s life ($550 * 18 years) there will be an additional ~$600K benefit by age 60 because of the pre-tax contribution benefit.

*If you use the 10.5% annual return that the Trump Account website uses then the tax saving alone would grow to $1.9M while each child’s account would grow to $18M by the time they are 60. Very generous assumption, but still…power of compounding is incredible.

Obviously if your employer is willing to just fund employees’ children’s Trump Accounts then that is even better for employees, but many companies won’t want to pay an additional $2,500/employee so this option is a major advantage for employees versus those that do nothing.

Process & Cost to Companies?

CFO: We are trying to cut costs so we can’t add more employee benefits like this

Wrong. Companies can add the program for just a little incremental cost that still has a large tax advantage for the employee.

The Department of Labor confirmed last month that Trump Accounts should not be subject to ERISA. That’s a big deal. Company 401(k)s, for example, are subject to ERISA and have a lot more additional documentation, reporting, audits, internal communications, etc.

Below are the main requirements to get it set up:

So maybe a couple thousand dollars to create the written plan and amend your existing cafeteria plan. And then your benefits provider will charge an additional small per participant fee.

Fairly immaterial cost for a material potential benefit.

Other Important Notes

• Tax-deferred vs Tax-Free: These accounts are retirement savings vehicles (IRAs) so eventually kids will pay taxes on gains on withdrawals at their tax rate at the time of withdrawal

• Money belongs to kid: The money is the property of the kid when it goes into the account. Parents can’t redirect money to another child like a 529 plan

• Money is locked until the year the child turns 18: There is no hardship exemption or any way to access the money before the kid turns 18. Don’t contribute if you think you will need the money before then

• Employees need to set up the accounts themselves: Employees have to set up their kids’ accounts and then provide that info to the company in order to participate in the program. Setup is really easy though…

• Need to track contribution limits: Someone who changes jobs or receives contributions from multiple employers must ensure total contributions do not exceed the $2,500 annual per-employee limit.

• Employers Have Contribution Options: Examples include, $1,500 funded for a new baby or adoption, employer match up to $1,000, flat dollar contribution per year, etc.

Final Thoughts

A company with a Trump Account Contribution program is a significant benefit for employees. And it will be a big negative in recruiting if you don’t have one.

I have heard many people (and even some companies) say they won’t sign up because Trump’s name is attached to it… Let me be clear, that is stupid. It’s a great benefit that will hopefully teach kids the value of investing while also giving some great tax advantages.

It’s also pretty cool that Grandma, Grandpa, or a friend can easily contribute to a child’s retirement account (scan a QR code and send money).

Set up accounts for your child and tell your company that they should set up the program as well!

Footnotes:

• Sponsors: Email me at onlycfo@onlycfo.io for sponsorship opportunities.

• Subscribe and tell your friends

Subscribe now

*Nothing in this post should be considered tax, legal, or investment advice. Educational purposes only. Consult an investment and/or tax advisor.

This will have a pretty massive impact on a lot of tech companies.

1. Vendor Spend: If your HQ is in California, then chances are you are paying 0% in sales tax on all your software/AI spend. But starting next year, most software/AI spend will be at least 7.25% higher than you were expecting.

2. Company Invoices: Starting next year, your invoices to California customers (tech companies sell to a significant number of CA headquartered companies) will be at least 7.25% higher.

That’s a significant price increase for both you and your customers!

Does this apply to all software? And…AI token spend?

The change applies to “prewritten computer software” (ie SaaS) transferred electronically or accessed remotely. But it excludes several categories, including “custom computer software” and “digital infrastructure”. The bill defines “digital infrastructure” as certain cloud-based services that let users create, deploy, scale, or run their own software (eg AWS, Azure, etc).

You will want to confirm the category your SKUs fall into. Especially as we get more definitions from California.

This will capture your Claude/ChatGPT/etc spend that is currently at 0% if you are in California. So your fastest growing (and maybe your largest) vendor spend is getting at least 7.25% more expensive next year. Ouch

FP&A Team Takeaway:

Budgeting: Make sure to add these additional costs to your budgets. This might be a material surprise cost in 2027. Pro tip: keep reading for how you might be able to reduce this sales tax liability.

Churn Impact: You also need to consider if/how this will impact your customer churn forecasts. This additional expense may be the breaking point for some of your customers.

What I am doing: Create a list of all your customers that will be impacted. Make sure to communicate the change with sales and customer success. And then consider the impact of any planned price increases next year. Maybe you want to keep prices more flat for those customers next year so it decreases churn?

Accounting Team Takeaway:

Collecting Sales Tax: If you have any decent sales tax software then you shouldn’t have to do anything with one important caveat…. I know some companies that don’t turn on states that don’t collect sales tax so they don’t have to pay additional software and filing fees. Make sure California is turned on if you have nexus there!

Do I have Nexus? Build a good nexus tracking process. I was once at a company that only updated it quarterly and then we sent out a ~$250K invoice that shouldn’t have included sales tax but didn’t because we hadn’t refreshed our nexus yet. We ended up eating about $18K in sales tax because we didn’t want to go back to the customer once we discovered it. More on this below.

What You Should Know About Sales Tax

Do I have nexus?

Nexus: level of connection between a business and a taxing jurisdiction that triggers a sales tax collection obligation.

There are two ways a company can establish nexus:

1. Physical presence. Typically gets triggered when you hire an employee in the state or local jurisdiction. Contractors and large amounts of inventory can also trigger it.

2. Economic nexus. Triggered when a certain $ level of sales and/or number of transactions to customers in the state. Below is a great visual for these different thresholds

California’s economic nexus threshold is $500K of sales in the current or prior calendar year, with no transaction-count threshold. So if you are a software company then you will trip this quickly.

Sales vs Use Tax

There are two primary types of indirect taxes in the U.S. Both are taxes on the same thing, but differ on who is responsible for paying it:

• Sales Tax: Collected by and responsibility of the seller because of their nexus in the state.

• Use Tax: Responsibility of the customer. Seller doesn’t have nexus (or at least didn’t charge sales tax) but the goods/services are subject to sales/use tax.

Idea is you can’t circumvent the tax by using out-of-state sellers to avoid the tax.

So in theory, the customer should be paying the tax either way so it should be included in total expense forecasting.

Where is Software Taxable?

Today, there are over 20 states in the United States that tax software. Many countries also tax software and digital goods from the very first sale. States and jurisdictions are continually changing their stance (like California) so it’s important to stay up-to-date.

Below is a map of states that tax software (with California starting in 2027)

The Dirty Secret of Many Software Companies

Most tech companies put their California HQ address on all their software purchases. It’s true that they are HQ in California but a significant amount of usage happens outside of California (particularly remote companies). It was also very convenient since they paid $0 in sales taxes in California.

But now California is taxing software, we are going to see many companies update their contracts with vendors…

Companies are going to reach out to their vendors to update their address or they will provide a “multiple points of use” (MPU) certification which states that usage happens in multiple places so it shouldn’t all be taxed at one address (this shifts sales tax responsibility to the buyer).

What About International Sales?

Companies are selling internationally a lot sooner today. Other countries have other indirect taxes that are paid by the customer like the value-added tax (VAT) or Goods and Services Tax (GST).

170+ countries have a value-added tax (VAT) regime with more than 80 applying VAT to digital goods and services transactions.

To determine whether VAT applies on a sale of your product or service, there are three main factors to understand:

1. Your customer’s location

2. The taxability of your product in that country

3. Your customer’s VAT registration status.

Final Thoughts

California now taxing software will be a big deal for a lot of companies. Plan accordingly.

If you don’t have a good process for sales/use tax, then you could end up with a huge liability that can derail fundraising discussions, a potential acquisition, or an IPO.

When the seller has nexus, it is generally their responsibility to collect and pay sales tax. Going back to your customers 2 years later and asking for sales tax is very painful. Ask me how I know…

Don’t wait to figure out sales tax. It will be much more painful later.

Footnotes:

• Email me at onlycfo@onlycfo.io for sponsorship opportunities.

• Subscribe and tell your friends

Subscribe now

*Nothing in this post should be considered tax, legal, or investment advice. Educational purposes only. Consult a sales tax advisor.