A Challenge to SaaS Orthodoxy

A Challenge to SaaS Orthodoxy

Klarna, the Swedish fintech giant, is making waves by churning from industry-standard software like Salesforce and Workday in favor of building its own internal systems with AI.

After their success with AI customer support automation which manages 2/3 of their customer inquiries , Klarna is now doubling down on this strategy.

Klarna is betting AI-enabled software is the future of internal tools. The corollary : the overall cost of building internal software with AI is lower than buying off-the-shelf solutions.

What is the break-even point for this kind of financial decision?

Let’s make a hypothetical example of MongoDB :

Over the course of a decade, the software spend could easily exceed $100 million. With the cost of software production falling1 & the cost of data storage also decreasing,2 the break-even point for building internal software is likely lower than ever.

How good of a CRM could a software company build with a $10m annual budget & with AI? It’s the equivalent bet to funding a startup with a $20-30m Series A & a big design partner.

Technology is always commoditizing itself. Perhaps bespoke software will have the same impact in sales as in customer support. That would provide Klarna a sustainable competitive advantage over time.

It’s also a forcing function to require the organization to rethink their workflows in the age of AI. More than just changing software, burning the boats & forcing a company to reimagine workflows with a blank slate can be a powerful way to drive innovation.

However, this approach isn’t without risks. Building and maintaining complex systems requires significant engineering talent and ongoing investment. Many companies have built internal systems only to eventually buy commercial offerings later after incurring significant expense.

If Klarna succeeds, the market for enterprise software could be upended with a fundamentally different architecture : data lake -> AI -> bespoke software.


Microsoft and ServiceNow have both reported 50-70% increases in software productivity as a result of AI. Amazon saved $700m in refactoring code with AI.

The major clouds have cut all data egress fees and the migration of data storage to standard formats like Iceberg on S3 plus the cost reductions of data sets create a deflationary environment for data costs. Not to mention anything about the scale discounts afforded to the largest users of cloud infrastructure.

Carlos Balbin

Venture Capital in Web3/AI/Fintech | Financial Modelling and Valuations | Consulting

1 个月

I have seen models focused on Great Data Lake and customized solution

Josh Weckesser

CEO ex-Salesforce, Xactly, DocuSign || Sales + Alliances || AI Productivity Accelerators

1 个月

Klarna just the first to do this. Great point on the forcing function to recreate the workflow with AI which can create massive change vs. bolt on solutions to existing bad data sets in CRM and HRIS.

Arnau Ayerbe

building the future of sales — prev. AI at JP Morgan

1 个月

the most interesting part here is perhaps how updates are a part of the process you don't mow your lawn once, it has to be maintained - which is the same as software building internally will make require paying the tech-debt and rellocating focus from the core product revolut is the example to follow for klarna revolut shipped an internal hr platform which they've now also rolled out as a product

Justin Hoover

Assistant Vice President - Senior Programmer at AllianceBernstein

1 个月

If it is truly possible to in-house the software, then it must also be true that off-the-shelf products could leverage the same technology to lower costs or a competitor that knows how to leverage the tooling could under-cut dominant market participants with a much cheaper product by several orders of magnitude.

要查看或添加评论,请登录

社区洞察

其他会员也浏览了