Why AI Lets Ambitious Business Owners Skip the "Start Small" Trap

Why AI Lets Ambitious Business Owners Skip the "Start Small" Trap

What if the conventional wisdom telling you to start small, prove one narrow use case, and slowly expand is actually the wrong advice for certain kinds of problems? That is exactly the conclusion Daniel Kjellén, the founder who co-built and sold the open banking platform Tink to Visa for €1.8 billion in 2022, arrived at when he sat down to build his next company, Freda, an AI-native compliance startup that came out of stealth in September 2026.

Kjellén's argument, laid out in a first-person essay for Crunchbase News, is rooted in hard experience. Scaling Tink across Europe to a team of roughly 700 people, he watched compliance grind momentum to a halt. Small wording tweaks triggered months-long reviews. Products were delayed by months at a time, a punishing cost in a fast-moving industry. When he and co-founder Fredrik Hedberg went back to the drawing board after the Visa acquisition, they identified compliance as the problem they wanted to solve. But compliance, as Kjellén describes it, is a "Hydra problem": legal tracks regulations, a separate department oversees implementation, product teams make their own feature calls, and yet another team checks whether those features were compliant. Policies, standards, and regulations live in entirely different silos. Fix one piece, and the problem immediately grows in another direction.

That insight led Kjellén to coin the counterintuitive concept of a "maximum viable product." For problems where every element is deeply interconnected, launching a narrow MVP does not just slow you down. It accomplishes nothing, because the partial solution does not change the underlying reality at all. His team at Freda now numbers 22 people, and he argues that AI-powered language models are what make this kind of comprehensive, fast-shipping approach possible today, something that would have required hundreds of specialists and many more years to attempt a decade ago.

For small and mid-size business owners, this framework carries a direct lesson about where AI fits into their own growth strategies. The temptation when adopting AI tools is to apply them narrowly: automate one email sequence, summarize one report, generate one batch of social captions. That is a perfectly reasonable starting point. But Kjellén's experience points to a more important question: are you using AI to chip away at the edges of a problem, or are you letting it change the structure of how your business operates? For many SMBs, the answer is the former, and the impact stays modest as a result.

The deeper takeaway is about the relationship between experience, capital, and ambition. Kjellén notes that his first exit gave him access to talent, partners, and investors that simply were not available when Tink was young. He could also fund the full build himself, which meant no investor pressured him to shrink the idea before it had been tested. Most small business owners are not sitting on a billion-dollar exit, but the structural principle still applies: the more you have proven about your own business model, the more aggressively you can invest in transforming it with AI, rather than just incrementally improving it. Credibility compounds, and so does the courage to pursue a bigger vision.

The Stockholm startup scene Kjellén describes, where Spotify and Klarna alumni routinely go on to found and back more ambitious companies, is a model worth studying. A culture of experienced operators who are willing to attempt harder problems, backed by AI tools that dramatically reduce the team size and time required to build comprehensive solutions, is a genuine competitive shift. For SMBs, the equivalent is building internal fluency with AI not as a one-off experiment but as a sustained capability, one that makes increasingly large problems addressable with a lean team.

Your actionable move this week: identify one operational problem in your business that you have been addressing only partially because a full solution seemed too complex. Map out every department or workflow the problem actually touches, using Kjellén's compliance example as a template. Then research whether an AI tool or combination of tools exists that addresses the full system, not just one node. You may find that what felt like a ten-person project is now a two-person project with the right AI stack in place.

AI is not just a way to do the same things faster. Used at the right scale, it is a way to tackle the problems you previously had to set aside entirely, and that is where the real growth advantage lives.

Originally inspired by: How I Learned That Some Startup Problems Are Too Big For A Minimum Viable Product (https://news.crunchbase.com/startups/minimum-viable-product-problems-ai-fintech-kjellen-freda/) See how Leads to Conversion can help you build an AI-powered growth strategy that goes beyond quick fixes. Lets chat

Your turn

What is your traffic actually doing?

Send us your details and we will come back with a short, specific read on what your traffic, your pages and your pipeline are doing today — and the first three things we would change. A real person reads every submission, and you get the read whether or not we ever work together.