I built an AI compliance tool and cold emailed 80+ fintechs. Here's what actually happened.
2 months.
80+ cold emails to compliance officers across Singapore, Indonesia, Germany, and the UK.
0 paying customers.
Here's the honest breakdown — what I sent, what happened, and what I'm changing.
What I was doing wrong
My CTA was "watch a 3-minute demo."
A compliance officer watches, thinks "interesting," closes the tab, forgets by Tuesday.
That's not a commitment. That's a view count.
My targets were CCOs at 200–500 person fintechs.
A CCO at a large fintech needs procurement approval, a developer, and three internal meetings before anything happens.
I was optimizing for the hardest possible sales cycle.
What I'm changing
The CTA is now:
"Integrate free in 30 minutes. First 1,000 decisions on us. No credit card."
A company that has pasted in an API key and sealed one real decision is a customer in waiting.
A company that watched a demo video is a cold lead.
The target mix is shifting too.
CTOs and developers at sub-50-person fintechs can paste an API call and be running before lunch.
That's the wedge.
The one metric I'm tracking
Free tier integrations.
Not replies.
Not demo views.
Not LinkedIn likes.
The number of companies that have actually sealed a real decision.
Everything else is noise.
Also changed how I'm explaining the problem.
I recorded a short video this week on the difference between logs and proof:
https://youtu.be/DCN6TU1NsiA?si=pDeBdlBLnuYOB0my
If you're building B2B and have figured out how to get the first 5 customers to actually use the product, I'd genuinely like to know what worked.
The point about targeting smaller companies really resonates with me.
I’ve noticed something similar: getting someone interested isn’t necessarily the hardest part — the number of decisions required before they can actually buy can be the bigger obstacle.
I’m experimenting with a niche product right now, and direct conversations have worked better than broad promotion so far. I think reducing friction after that initial interest may be just as important as finding more leads
Anthony, the shift from optimizing for demo views to getting someone to actually integrate sounds like the right move.
One thing I'd be curious about is what the 80+ non-buyers had in common. There might be a useful pattern in who replied, who showed interest but stopped, and who never engaged at all.
Have you looked at the responses from those 80+ companies as a group yet?