Becoming AI native took a while. The whole journey started about ten years ago, when I was shutting down my startup because I was in a market that was too small, and too early on top of that.
And because I had promised my wife I would never do another startup, I went to work as a consultant, passing on what I had learned in international sales, marketing and project management. With CEOs I soon started to see that they all make the same mistakes and have the same problems, just at a different stage of growth. That led me to building companies systematically, as a glass box, and to a validation framework that is now called Compass. Turning it into a product in 2020 was too complicated, though, so I put it on ice for a while.
Then ChatGPT 3.5 came along and it was clear to me that this was something I would really get along with. I've been hooked ever since. Last year I finally quit a well-paid job as a sales director and went off to build my own AI startup.
Can consulting scale?
The topic is clear to me: scaling consulting. Can you turn consulting into a product and really scale it? My first brief was bold: there would be no app, the CEO would just talk to AI and it would build the whole company for them according to the framework.
Six months later I found out it wasn't going to work that way, and it was my mistake in the brief, not the people who worked on it. I had to let go of the idea that AI would somehow magically solve it and build it on its own. That's where I went from AI first to AI last, but still AI native.
AI on a short leash
I then took development closer to me and built it differently. Every change has a recorded decision on why it was made, goes through tests and an independent check before it reaches a client, and we only ever read client data, we never change it. AI doesn't decide anything on its own at Smartin; it works inside rules that a human holds. My favourite word when it comes to a B2B application is determinism: same data, same answer, every Monday.
And lo and behold, it started to work. Once a builder, always a builder, and the joy of getting done in an afternoon, thanks to those rules, what used to take a week is indescribable.
No longer the whole company, just the customer journey
Along the way I learned one more thing: when you try to fix the whole company, you end up fixing nothing properly. So we narrowed it down. Today Smartin deals only with the customer journey, from the first website visit through sales to customer care and contract renewal. We look for where a company is losing money along it. It often looks like sales is stuck, and the problem is somewhere else entirely: in marketing, in the offer, or in who you sell to. Every Monday the CEO gets one thing to do, and six weeks later we measure whether it helped. If it didn't, we stop doing it.
I have my first clients we are testing Smartin with, and we can see it really delivers results. We are now moving from a private beta to a public one. I have no ambition to be a solo billionaire entrepreneur forever, but who joins the team and what they will do has changed radically since the first concept. So I'm curious where it will all take me.
Compass for free
Compass, where it all started, I give away for free. If a tool for validating ideas would help you, you can try it here.