You share your product in a Slack group. You post it on Twitter. You send it to three friends.
Most of them never open the link.
The ones who do spend 90 seconds on it and close the tab. Maybe they mean to come back. They don't.
And the rare few who actually try it? They send back: "Really cool!" "Love the idea!" "This could be huge!"
Nobody tells you that your onboarding is confusing. Nobody mentions they had no idea what to type first. Nobody says they bailed after 40 seconds because they couldn't figure out what the product actually does.
They're not lying. They just don't want to be the one to tell you. And honestly — you didn't make it easy for them to tell you, either.
---
I've been building indie products for a while, and this kept happening to me. Three separate problems, every single time:
Getting anyone to test at all. Most people say yes and ghost. The ones who do try it are doing you a favour — which means they're not going to be harsh.
Getting them to go deep enough. A 90-second skim isn't a test. You need someone who actually tries to use the thing, hits a wall, and notices where the wall is.
Getting feedback you can do something with. "The onboarding felt a bit confusing" is not actionable. "I didn't know what to put in the first field and gave up" is.
The root cause of all three: there's no structure, no incentive, and no skin in the game.
---
So I started thinking about what would fix all three at once.
Incentive: make it a swap. You get feedback because you give feedback. Everyone in the loop is a builder who wants the same thing — so they actually show up.
Depth: require them to use the product before answering. Not skim it. Use it.
Actionability: replace "any thoughts?" with sentence completions that are almost impossible to answer vaguely:
- This product is useful when I…
- I got confused when…
- The person I'd send this to is…
- I'd pay for this if…
- If I had a magic wand, I'd want…
- This reminds me of X, but…
---
That's TestPilot.
A reciprocal feedback swap for builders. Share your product, get six structured, specific reviews from people who actually used it. Spend ten minutes giving the same honest read on someone else's. No panels of strangers, no credits, no "looks great" non-answers.
Every round gets distilled into a punch list — what came up most, and the single sharpest thing to fix next.
We're in alpha, waitlist is open now.
What's the worst "feedback" you've ever gotten on something you built? And what would actually useful feedback have looked like?
→ [testpilot-alpha.netlify.app](https://testpilot-alpha.netlify.app)
The shift from "collecting opinions" to "designing a process that produces actionable evidence" was the part that stood out.
Most feedback systems fail because they optimize for getting responses, not for generating observations someone can confidently act on.
Yes — and that's exactly why the prompts matter. "What do you think?" produces opinions. "I got confused when…" produces evidence. The structure does the work so you don't have to dig for the insight afterwards.
Appreciate the context.
The difference between collecting responses and creating actionable evidence is the interesting part here.
Would be good to understand more about how you're seeing users apply those insights.
What's the best email to reach you on?