I launched a site to see if people would pay for your project — and you helped me shape it

the two-audience split isn't the real fix though, the vote itself is. "yes I'd pay" costs nothing to click so a doctor and a developer voting yes look identical in your data even after you fix distribution. i lean on viewfy for surfacing threads where someone's already complaining about the problem, not asking who'd pay for the solution. worth weighting votes by whether the voter describes having the problem, not just their tag.

The audience-mismatch problem you spotted is the whole ballgame for a validation tool. A "yes I'd pay" from the wrong crowd is worse than no signal, because it's falsely reassuring. Two thoughts. Let idea-posters tag the target audience and show voters that context, then segment or weight results by whether the voter is actually in that audience. And keep in mind "would you pay" is a weak signal even from the right people, since stated intent and real payment diverge hard. A stronger version asks voters to do something with a small cost, like leaving an email to be notified at launch or picking the price they'd expect, so you're measuring revealed interest rather than politeness.

the audience mismatch thing you found is real, but i'd be careful about "fix it by getting a broader audience". broad voting on a doctor-facing idea just gives you louder noise. what actually moved the needle for me on validation was ignoring votes entirely and looking at whether the person voting has the problem today. two cheap things you could ship without rebuilding: 1) make the poster pick who the idea is for (role + where those people hang out), and show votes split by "i am that person" vs "i'm just an indie hacker" — one number is signal, the other is vibes. 2) replace "would you pay" with "have you paid for something like this in the last year, and how much". intent-to-pay is nearly free to answer yes to; past spend isn't. also fwiw the reddit plan works better if you don't post the platform at all — go answer questions in the niche sub for a couple weeks first, then the invite lands as a person asking rather than a founder farming votes.

Audience mismatch is actually a measurement collapse problem. IdeaPz is currently measuring "would tech people pay for this idea" not "would the target audience pay for this idea." These look like the same measurement until they diverge.

The subtle insight: "idea validation" hides three different measurements. First, signal strength (the idea has a problem worth solving). Second, audience-product fit (the target audience has this specific problem). Third, willingness-to-pay precision (they'd pay THIS much, from THIS audience segment, at THIS stage of awareness). When you measure all three at once, you can't tell which broke.

Your two-audience direction solves for this. But measurement design matters: if 1000 developers upvote a "doctor scheduling" idea, that's not weak signal - it's a clear signal that developers aren't the customer. That's valuable information you'd miss if you only counted votes. The metric that matters is the mismatch itself: how much upvoting comes from outside the target audience, and do those outsider votes correlate with zero email signups from the actual target audience?

One measurement gap worth closing: willingness-to-pay without friction. A doctor might say "yes I'd pay" on an anonymous voting site, then ghost when you ask for email. That's not a validation failure, it's a measurement boundary problem - you're measuring interest, not commitment. Measuring commitment requires friction (asking for email, capturing actual intent signal) which might suppress the upvote count but increases the accuracy of your "would actually pay" signal.

The real play: publish how much your metrics diverge between audiences. Show creators "1200 tech people upvoted this, but 50 doctors did" - that divergence IS the validation data. Most validation tools collapse it into a single score.

I really relate to this, especially the audience-mismatch part. One thing I’ve learned while building is that getting feedback from the “right” people is often more valuable than getting a lot of feedback from people who simply think the idea sounds good.

I’ve been trying to apply the same thinking with Pordix — instead of building a huge all-in-one platform, we’re focusing on practical problems businesses already have, like SEO content, website performance, and getting discovered through AI search.

It’s still very much a learning process for us too, but building around an actual pain point rather than what sounds cool has made product decisions much easier.

For anyone curious about what I mean, I’m building it here: https://pordix.com/

I’m also really interested to see how IdeaPz evolves once you start getting feedback from people who are actually the target users rather than just other builders.

The audience-mismatch fix is the right call, and I would push it one step further based on what we see validating outreach offers: votes measure curiosity, only pre-orders measure intent. If you want IdeaPz to produce decisions rather than good feelings, let a creator mark a project as "committing" - voters who say yes get asked for a 5-dollar refundable reservation or just an email plus a stated budget. The yes-but-wont-pay crowd sorts itself out and the creator learns what a buyer looks like, not what a voter looks like.

Also, the email capture moment matters: collect the address right after the yes vote while the justification is fresh ("leave your email and tell the maker why you said yes"), not in a separate step. Those reason-notes become the creator's first marketing copy - you would be sitting on the exact words buyers use, which is worth more than the vote count itself.

Bookmarking this - happy to stress-test the validation angle on the next batch of projects.

The main problem is the audience mismatch. It's not a small calibration issue. When a developer says they'd pay for a tool built for dentists, they're mostly judging whether the idea sounds plausible. They aren't telling you whether dentists actually struggle with that workflow. I'd make every project name its target role and use case. Then I'd treat votes from other audiences as interest, not validation. Five short interviews with people who do that job will teach you more than 100 votes from a general audience.

The audience-mismatch point is the one that'd worry me most as a builder — a yes/no vote from someone who was never going to be your buyer isn't validation, it's noise that feels like validation, which might be worse than no signal at all.

I'm mid-validation myself right now (landing page + waitlist for a fairly niche small-business tool), and the price-point-instead-of-yes/no suggestion in the comments is something I'm stealing immediately — "would you pay" and "would you pay $9/month" get you very different answers, and I'd been lazily doing the first one.

On the two-sided pivot: my instinct is that trying to serve both creators and problem-solvers at once is exactly the trap the audience-mismatch feedback was warning you about, just at the platform level instead of the individual-idea level. What made you lean toward two-sided instead of picking one side and going deep, like a few commenters suggested?

The gap between "I'd pay" and actually paying, that others mentioned here, hits close to home for me, but from the other side.

I just launched something I built solo and priced it low since I have zero sales so far — no proof it works, just a bet.

On the two-audience question ( creators + users who discover ideas ) — two concrete things that might help:

1. Turn the emails you already collect into a real audience, not just a list.
   When a creator gets emails from people interested in their idea, offer to send them one single follow-up question: " would you be up for testing this before the public launch ? " ... Whoever says yes becomes a real test audience for that specific creator, without you needing to bring general traffic to IdeaPz itself.

2. Build a dedicated page per target audience instead of one single page.
   If a creator tags their idea "doctors" or "contractors", auto-generate a page like ideapz.com/for/doctors that the creator can share directly in their own circle — that replaces hoping random users stumble onto it while browsing, and gives you
   Google-indexable pages as a side effect too

I think you already found the real problem yourself. I ran two versions of this, a page with a promise and an email box, zero emails on both. A "yes, I'd pay" vote is pretty cheap. Have any of your votes so far come from someone who actually has the problem, or is it mostly builders?"

I like the idea because it tests willingness to pay before someone spends weeks building. One thing I’d be curious about is how you separate genuine buying intent from people who simply like an idea, and whether you could add a stronger signal—like joining a waitlist with an expected price range. could make the validation even more useful.

Awesome, helps me out a lot with similar topic!

Cool direction - the audience mismatch problem makes sense, that's basically the same reason I've been hesitant to trust "just post it and see" validation methods for niche ideas. A dev-heavy crowd voting on whether they'd pay for something aimed at, say, small business owners or doctors isn't really validation, it's a proxy at best.

On attracting the "user" side specifically - I'd guess the hard part isn't traffic, it's motivation. Indie Hackers users show up wanting to build/validate their own stuff, so they've got a reason to engage. Randoms outside tech don't have that built-in incentive to visit a validation site unless they're already annoyed by the specific problem being tested. Might be worth leaning into niche communities/subreddits per-project rather than trying to build one central "user" audience - e.g. redirect voters from wherever the actual target audience already hangs out, rather than hoping they come to IdeaPz directly.

Following this thread with interest since I'm mid-way through validating my own idea right now - curious how the two-sided thing plays out.

the two-sided thing is the hard part. you basically need both sides to show up before either side sees value, which is the classic chicken and egg problem. i'd think about seeding the users side manually at first, go find people in niche communities (reddit, discord, slack groups) who match a specific project's target audience and personally invite them to vote on that one idea rather than trying to build general traffic to the platform. easier to prove it works for one vertical before trying to scale both sides at once

On attracting both sides, do not. Make the creator bring the voters: give every idea a shareable link and a simple embed so the founder can drop it into the community where their doctors or accountants already hang out, and your job becomes hosting the vote instead of sourcing the audience. That fixes the mismatch without you needing traffic from a second market you have no way to reach yet.

Useful shape, but the honest warning: "would you pay for this" votes and actual payment are two different signals, and the gap is usually 20:1. People vote yes because the idea sounds reasonable, not because they have the problem this week.

If you want the votes to mean something, force the voter to pick a number and a frequency — "$19 one-time" vs "$9/mo" vs "no" — instead of a thumbs up. The number is where the polite yeses fall apart, and it also tells the poster whether they're selling a one-time artifact or a subscription, which is a totally different product.

Second thing that would help: ask the voter whether they've paid for anything in that category before. A yes from someone who already bought a competitor is worth ten from someone who hasn't.

The mismatch you called out is the product. Votes from the wrong room are worse than no votes — they feel like signal and steer you into building for people who will never pay.

I'd pick one dense buyer community that matches a single niche of ideas on the board (one specialty forum, one professional Slack/Discord, one Facebook group where the buyers already hang out) and stay there until that niche produces returning posters and returning voters. Don't broaden IH traffic until one non-IH room works.

Metric to watch: returning users on both sides in a week, not installs or total votes. If creators post once and vanish, or voters never come back, the loop isn't real yet.

Freeze feature work (share buttons, discovery polish, new surfaces) until someone in that niche pays or at least returns without you nudging them. Two-sided cold-start is usually "one niche deep," not "both sides wide."

The audience mismatch feels like the more important discovery than the traffic problem.

Curious whether you’re thinking of solving this by matching projects to relevant user groups, or by building a separate user-side discovery loop that naturally brings the right people in.

The audience split is a useful distinction. I’d treat “would you pay?” as a weak signal and track a sequence instead: whether the voter is the target user, whether they have the problem now, what workaround they use, and whether they have paid for a solution before.

A smaller sample of people who currently feel the problem can be more informative than a large general vote. Reporting conversion by audience segment, rather than only one aggregate number, should make that signal easier to interpret.