Launching v3 in a few days after v2 got 8 upvotes. Before I do, I want the harsh version of your feedback.
v2 of our product launched to a resounding 8 upvotes. I've had a while to sit with that, and honestly it was the most useful failure we've had.
The sting wasn't the number. It was realizing why. v2 did one thing validate a startup idea against real market data instead of vibes. People would try it, nod, and leave. The verdict was correct and completely inert. Nobody wakes up wanting "your idea scores 7/10." They want the next step, and we handed them a dead end.
So v3 is basically us admitting that. We kept the validation, but now it doesn't stop at a verdict it carries through into a PRD, information architecture, wireframes, and an actual working prototype. The bet is that the build should start from what the market data says, not from whatever you happened to type into a prompt.
The part I'm proud of: the first screens aren't a guess. They're shaped by what users of existing products are actually complaining about. In my head that's the whole point most "idea to app" tools are confidently generating the next plausible screen with no idea whether anyone needs it.
The part that keeps me up: "idea → working app" is Replit, Lovable, Stitch territory. Rooms full of teams with 100x our funding. I genuinely can't tell anymore whether "the build starts from validated data" is a real, defensible difference or a comforting sentence I wrote so I'd feel okay walking into that fight.
A couple of small things I already changed from the harsh v2 lessons: I cut the celebratory "your idea is validated!" moment entirely turns out a pat on the back is exactly what made people leave. And I stopped leading with the data volume ("200k+ data points") because nobody outside the building cares how hard the plumbing was.
So before I launch, one honest question: if you landed on a tool that promised "idea → validated → working prototype," what would make you actually trust it enough to keep going instead of nodding and closing the tab like my v2 users did?
8 upvotes is interesting data, but upvotes != demand. The hard question is: did any of those 8 try to pay or ask when they could use it? Polite feedback is abundant, willingness to pay is scarce. I'd find the 1-2 people who engaged most and offer them early access in exchange for a real conversation about what they'd actually pay.
yeah, none of them asked when they could pay, which is the actual answer. polite nodding was abundant, that's the whole story of v2 honestly.