Build vs Buy | The End of SaaS?

Klarna makes headlines for ripping out Salesforce and Workday and replacing with AI. Are companies going to build all their SaaS tools internally?

Sponsored by Brex. Control spend, earn rewards, and automate accounting with Brex.

Why do we still pay vendors and suppliers by check, ACH, or wire? Checks are prone to fraud, ACH payments are slow, and wire transfers are costly. Modern finance leaders prefer to pay vendors using corporate cards. Our guide, “The future of B2B payments,” tells you the reasoning and why this small change can deliver big returns.

---

Klarna Replaces SaaS Tools with AI?

There is a lot of doomerism and hype on social media in the software industry when there is a new AI development. It’s important to remember that social media can be an echo chamber and some things get blown out of proportion. This may come as a shock to some, but just because it’s on social media doesn’t make it true (at least to the degree it appears).

To be clear, AI will disrupt a lot of companies (many SaaS companies particularly). BUT…this will not happen overnight and most SaaS companies are not going to roll over and die like many are anticipating. Many will adapt, but there will also be many that die.

The below remarks from Klarna’s CEO made the rounds on social media this past week:

The announcement was met with two extreme sides:

1. This is the END of SaaS

2. Skepticism, caveats, denial, etc.

The first group was quick to claim that this is a warning signal that all of SaaS is going to $0 — “If a company can replace Salesforce then all SaaS can be built in-house!”

The second group is a bit more skeptical and denied what Klarna did was possible. There is likely some bias in this group because maybe they are involved in the software industry and want/need it to flourish.

I fall somewhere in the middle. My belief is that Klarna’s claims aren’t as good as they sound but there is some truth behind them. And we will start to see a lot more of this in the next 1-2 years.

Claims of Ripping out Salesforce and Workday

Klarna just fully ripping out these two systems (*and it being successful*) would be a REALLY big deal for the future of cloud companies because these two companies are thought of having giant moats. They are systems of record with lots of complex workflows. So if AI makes it easy to rip them out then is any SaaS company safe?

Here are some of my initial questions after hearing this Klarna announcement:

• Can Klarna really just use some AI to replicate everything Workday and Salesforce does?

• What happens with all the integrations?

• What about payroll compliance?

• How can they go public without a proper CRM? *they are private but eyeing an IPO soon

• Is it really worth to build all of this in-house? They are taking away in-house engineering time

• etc etc etc

The full claims didn’t really pass the sniff test. While there for sure is *some* truth to the AI efficiencies, just ripping these two tools without replacing them with something (which is what some were led to believe) would be really hard.

So what did Klarna actually do?

This isn’t the first time that Klarna has made really big claims about AI. Klarna’s CEO seems to have been making a lot of AI efficiency claims recently. Below are a few:

• 50% reduction to customer support because of AI

• Klarna’s AI assistant is performing the work of 700 employees

• Klarna plans to reduce its workforce from 3,800 to 2,000 because of AI

• CEO boasts about being an OpenAI guinea pig

I am not saying there isn’t some truth to the claims because AI is definitely making Klarna and most other companies more efficient. But Klarna might also be pointing mostly to AI when there is more to it. Also, the free marketing coming from all these AI announcements leading up to a Klarna IPO is probably nice :). Being seen as a leader in AI might be good for their valuation…

Killing Salesforce

Several people on social media pointed out that Klarna appears to just be spinning a decision Klarna made years ago to churn off of Salesforce. Moving off because of AI sounds a lot cooler. Below is from a Freshworks transcript in 2022:

source: Twitter

And almost exactly two years from the above transcript Klarna is churning from Salesforce…

Ripping Workday

Ripping Workday is a similar story. The vast majority of companies would never want to build a payroll tool in-house. Even with AI it is way too risky and not worth it.

This is not what Klarna is doing.

Klarna has been using Deel for payroll and most of HR for a while so I am not sure what they were really even using Workday for by the time they churned. Klarna’s CEO talked about the move to Deel last month.

Impact of AI

Klarna is certainly saving money by making these switches (simpler and less expensive tools). Part of the savings may come from certain things they can now do with AI. And the other part is just a simplification of their business processes. Salesforce and Workday can handle a lot of complexity and if you don’t need all the extra stuff then money can certainly be saved through simplification.

I am not saying to ignore the Klarna claims because claims like this will become more common and more real. But…be careful of how much you read into them.

• Klarna wants to be seen as a leader in AI (and they probably are to some degree)

• Klarna wants the free marketing from all the buzz

• Klarna wants to build excitement around the company before an IPO and an AI leader probably fetches a higher valuation

• Perhaps (like many companies) Klarna dramatically over hired and overspent in 2021 and is using the AI story to justify why it is cutting. Sounds a lot better than admitting to bad decisions.

While completely ripping out major SaaS tools is still difficult today, we will start seeing more of this in the coming months and years (and it will be truer).

Build vs Buy

The build vs buy debate is nothing new. Companies, especially larger ones, frequently consider building some of their tech in-house. For a long time, the answer usually was to buy the tech because building was too hard.

AI is rapidly decreasing the time and cost to build internally, so a lot of people are quick to conclude that there will be mass churn from all SaaS vendors as companies build out their own tools with AI. “Prompt ChatGPT they you want a CRM like Salesforce and boom!…you got one”.

While development costs/time are rapidly shrinking, this assumes that the creation of the product is all that people are buying….

There is a lot more thought that should go into the build vs buy analysis:

• The maintenance of the product

• Support of the product

• The opportunity cost of working on something not core to what you sell

• Thought leadership and knowing what to build

• The data moat that vendors have. Increasingly critical with AI use-cases.

The tech debt (extra work required to fix a product because of shortcuts and band-aid solutions) and people costs build up over time as a company scales. Yes, it may be relatively easy to spin up a simple CRM now, but what happens in 6 months when there are 200 internal requests, integration issues, scalability problems, etc. Tech debt builds and it becomes a much bigger time suck for your internal team.

To be clear, the decision to build instead of buy may start to make more sense in a lot of cases now (and will continue in the future). BUT…just because a company can build it doesn’t mean they should.

The below comments about large corporations’ adoption of AI to build out internal tools is generally correct as well. It will start to happen but it certainly will not happen overnight. And in the meantime, the smart SaaS companies will adapt to the new environment.

SaaS Is Not Dead (yet)!

SaaS companies aren’t going to just roll over and die, but many won’t adjust fast enough.

There will likely be a lot more companies who choose to build versus buy in the future. They will go in assuming AI will make it easy and they can save a lot of money. Sometimes they will be right and they will save some money. Other times they will be terribly wrong and waste resources and prioritize the wrong thing for their company.

I am all about saving money…but make sure building internally is really worth it. Sometimes it will make sense to build, but a lot of the time it still won’t make sense (yet…).

---

Footnotes:

• Check out Brex’s guide: “The future of B2B payments,”

• Reply to this email if you need a fractional CFO and/or bookkeeping for your company. I partner with a great group.

• Check out OnlyExperts for to find offshore accounting resources. They have some amazing talent for 20% the cost of a U.S. hire.