I've been running 3 MVPs this year and finally figured out how to stop building features nobody wants

Totally relate. Even just talking about my idea with friends was enough to reveal a problem. Their first reactions made it obvious I was obsessing over features nobody actually wanted. So I guess it's a useful first reality check before building too much.

This really hits home. The dashboard story is the perfect example — we all have features that feel obvious until you realize nobody actually asked for them.

The thing that changed my thinking was separating 'this sounds like a good idea' from 'people keep running into this without being asked.' The second one is the only signal that really matters early on.

Going through something similar right now with my own project — the hardest part is not building what excites you but what keeps showing up in conversations you didn't start.

Have you found that users are more honest in their feedback when they don't know you're tracking it?

This really resonates — especially the part about checking actual patterns instead of building based on assumptions.

One thing I didn’t expect in my own experience is how different “interest” vs “intent” really are. I had a few hundred followers pre-launch for a project, but converting that into actual users/backers has been much tougher than expected.

It’s made me rethink what signals actually matter early on — not just what people say, but what they actually do.

Curious — have you found any specific signals that consistently indicate real intent vs just passive interest?

This really resonates — especially the part about thinking something is “obvious” and then getting zero response after shipping it.

One thing I’ve been noticing is that the problem isn’t just tracking feedback, it’s that a lot of the most valuable signals never show up as structured feedback in the first place.

People don’t always submit feature requests — they complain in Reddit threads, ask vague questions, compare options, or drop little frustrations in comments.

In the space I’ve been looking at, the signal is everywhere, but it’s fragmented and hard to interpret. You end up either overbuilding based on noise or missing patterns that are actually there.

Curious — have you found that most of your “useful” insights come from direct feedback, or from patterns you had to piece together yourself?

This is a great example of why building faster doesn’t automatically mean building better.

AI has made it ridiculously easy to ship features, but it hasn’t improved judgment. If anything, it makes it easier to go down the wrong path faster.

The “no one complained about it in 5 months” point is gold. Most of us optimize for what’s visible (what we can design/build), not what’s actually painful.

Feels like the real skill now is:
pattern recognition > execution speed

Curious if you’ve noticed certain types of feedback being more reliable than others?

Building right now and have been dragging it out for longer than I ever expected. Only recently did we actually start getting real users and understanding what they wanted. I sometimes think being reactive when you have only a few users is also a distraction and that sometimes you do have to follow your gut but it is definitely a balancing act for sure.

Yeah, this is such a common trap — building what feels right vs what users actually keep asking for. The shift from scattered feedback to pattern-based decisions is honestly a big unlock.

Even we’ve been trying to stay closer to real signals instead of assumptions — especially early on when every feature feels important but not everything is actually needed.

Also sharing something I’m building in parallel — You have an idea. $19 puts it in a real competition. Winner gets a Tokyo trip (flights + hotel booked, minimum $500 guaranteed). Round just opened, so best odds right now: tokyolore.com

Ziraxo — Create AI-powered 3D images and models in seconds! Easy, fast, professional. Ziraxo.com

This really resonates. One of the hardest parts of building is separating a real repeated pain point from something that just sounds like a good idea. I’m currently at a similar stage with an AI product, and I’m realizing that pattern recognition in feedback matters much more than intuition.

We learned this the expensive way. Built a landing page auditor, got 8 signups, zero revenue. The tool worked perfectly. The audience didn't care enough to pay. What changed everything was going through 530+ competitor reviews and mapping the actual complaints by category. The patterns that emerged were completely different from what we assumed users wanted. Structured feedback is valuable but the best signal comes from what people complain about when they think nobody is building a solution.

Does this work as intended? Because, as it seems to me, such a module can now be created very quickly and efficiently directly in your application using Claude, Cursor, etc., and to be independent of external services.

the feedback loop was the real fix, not the tool itself. spent weeks building PM dashboard features nobody asked for before I started actually tracking what people complained about. the tool just makes you do the discipline.

This is such a big shift — moving from “idea-driven” building to “signal-driven” building.

I think a lot of us fall into that trap because building feels like progress, but without structured feedback it’s basically just well-informed guessing.

What you described about killing the dashboard redesign is key — not all good ideas are valuable problems. The absence of complaints is often a stronger signal than excitement.

I’ve been thinking about this more as a feedback density problem — one-off requests vs recurring patterns. The real leverage comes from spotting what keeps showing up unprompted.

Curious — have you found a good way to capture implicit feedback too? (like user behavior or drop-offs, not just what people say)

This is such a real shift.I’ve been going through something similar — realizing most of what I build is based on assumptions, not actual signals.Recently started experimenting with a different angle on the “finding users” side — pulling YC startups that are actively hiring and reaching out to founders there.Still early (only ~27 leads so far), but trying to see if catching people at the right moment changes how they respond.Curious — how are you getting your initial users right now?

Pretty cool idea, I was eager to try but google auth didnt let me in.