We almost built a product on an assumption that turned out to be false
Here’s a small moment from our build — one that almost certainly saved us a lot of wasted work — and it’s worth handing to you exactly as it happened.
We got excited about a product idea — a plug-and-play onboarding experience that walks a beginner into using AI, game-like, fun, where you build something real as you go. The kind of idea that makes you want to open a blank file and start building immediately. And the monetization felt clean: we’d earn through referrals to the AI model itself. Refer the platform someone already needs, get a cut, everybody wins. We even called it, in our own notes, “the cleanest monetization we’ve found.”
Then, before we built anything, we checked whether that referral program actually existed.
It mostly didn’t.
The five-minute check that changed the plan
The assumption was simple and sounded obviously true: the big AI models have affiliate programs, so we can earn by referring people to them. It’s the kind of thing you’d never think to question — of course they do, everyone has an affiliate program now.
So we spent a few minutes verifying it instead of assuming it. What we found:
- The major consumer AI assistants largely do not offer a standard affiliate program an individual can just sign up for. Some have enterprise-only partner programs; some run occasional limited referral offers that can be withdrawn at any time; the reliable, open, “sign up and earn recurring commission” version we’d pictured wasn’t there.
The single most important piece of our monetization plan — the part we’d literally labeled “the cleanest” — was largely not available to us. We learned that before writing a line of product code, instead of after.
Why this is the whole lesson
It would have been so easy to skip that check. The idea was good. The logic was clean. We were excited. Every condition that makes people build on sand was present: enthusiasm, a plausible assumption, and a reason to want it to be true.
The discipline that saved us is boring and it’s this: find the single assumption your plan depends on most, and verify that one before you build on it. Not all your assumptions — you’d never start anything. Just the load-bearing one. The plank the whole structure stands on. For us, that was “the monetization exists.” If that’s false, the elegant product on top of it doesn’t matter.
There’s a quiet trap in being excited about an idea: excitement makes you want to protect the idea, which means you start avoiding the questions that could kill it. The healthiest thing you can do for a plan you love is to point the hardest question straight at its foundation — early, while changing course is still cheap.
The good news: the idea survived
Here’s the part that makes this hopeful rather than deflating. Verifying the assumption didn’t kill the product — it killed the wrong monetization, while there was still time to find a right one. The onboarding idea is still strong. What changed is that we now go looking for monetization that actually exists: affiliate programs for the supporting tools the onboarding genuinely uses (the ones that do have open programs), or using it as a front door to our own offers. A real foundation, not an assumed one.
That’s the difference between finding out now and finding out after a month of building: now, it’s a redirect. Later, it’s a teardown.
How to use this yourself
If you’re about to build something — a product, a side project, a new offer, an automation — try this before you start:
- Name the load-bearing assumption. Finish the sentence: “This only works if ______ is true.” That blank is what to check first.
- Spend the cheapest possible effort verifying it. A few searches. One email to someone who’d know. Reading the actual terms instead of assuming them. Usually minutes, not days.
- Let it actually be allowed to fail. If you go in needing the answer to be yes, you’ll find a way to read it as yes. Verification only works if “no” is a real possible outcome.
- If it fails, redirect — don’t mourn. A false foundation found early is a gift. The idea on top of it is usually fine; it just needs a foundation that’s real.
The most expensive thing in any build isn’t the work. It’s building the right thing on the wrong assumption — and only finding out once it’s load-bearing. Five minutes of honest checking, up front, is the best deal in the whole process.
So before you pour the foundation, tap the ground. It’s a small, unglamorous habit, and it will save you more time than almost anything else you do.