4
27 Comments

2 days from Product Hunt launch — here are the real numbers so far

been building IdeaDose for the past few months — an AI tool that evaluates startup ideas and gives a GO/RISKY/KILL verdict. not the "great idea!" kind of validation. the honest kind.

launching on Product Hunt this Thursday (March 26). here's where things actually stand:

the numbers

  • 84 unique visitors last week (116 sessions)
  • 19 signups (12% conversion from landing)
  • traffic sources: 55% direct, 13% Threads, 13% X, 7% Google
  • 83 dead clicks on the landing page (ux issues i'm still fixing)
  • bounce rate spiked to 92% on one day — killed email verification, switched to oauth only

what i learned building this

  • tested 15+ competitor tools. most gave a terrible idea 78/100. nobody wants to be the bad guy
  • built kill criteria (K1-K5) as deterministic logic, not AI opinion — so the verdict is consistent
  • every claim in the report links back to real data (reddit threads, github stars, pricing pages)
  • cut 3 features last week that nobody asked for (team collab, pdf export, saved comparisons)

what i'm nervous about

  • "just ask chatgpt" is the real competitor, not the other validators
  • 84 visitors in a week is not a lot. distribution is the hard part
  • first-time PH launcher. no idea what to expect

if you've launched on PH before — what actually moved the needle on launch day? genuine comments, social posts, DMs to friends?

ideadose.dev if you want to check it out. will report back with PH results next week regardless of how it goes.

on March 23, 2026
  1. 1

    update: launched on PH yesterday.

    promised to come back with the results, so here they are:

    • 30 visitors on launch day (8 from PH)
    • 2 signups
    • 2 evaluations requested, 1 completed
    • not featured
    • organic PH comments from strangers: 0

    the one bright spot: users who tried it spent 9+ minutes reading reports that took 47 seconds to generate. so the product holds attention — distribution is the problem.

    biggest lesson: 2 weeks of community building wasn't enough. everyone here who said 2-6 months was right.

    thanks for the honest feedback on the D-2 post — especially @GrolyDev @microbuilderco @NsimbiniSol on the outreach timing. wish i'd started earlier.

    next step: stop treating PH as THE channel and diversify.

    1. 1

      Proud of the progress you're making!

      1. 1

        thanks! the numbers are small but at least the product works when people find it. distribution is the next puzzle to figure out.

  2. 1

    Those pre-launch numbers are solid — the email list size matters less than the activation rate, and 23% open rate with 8% CTR tells me your messaging is actually resonating.

    A few things that helped our PH launch go better than expected:

    1. Personal outreach to your existing users the morning of launch — not a blast email, but direct messages asking them to check it out if they found value. Those early upvotes in the first 2 hours set the trajectory.
    2. Reply to every single comment on PH within the first 4 hours — the algorithm rewards comment velocity and it shows product-minded founders behind the product.
    3. Do NOT post in Slack/Discord groups asking for upvotes (PH can detect this and will suppress your ranking).

    What time zone are you launching in? The 12:01am PST timing is conventional wisdom but I have seen afternoon launches work too when your audience is non-US.

    1. 1

      appreciate the detailed breakdown — taking notes on all of this.

      personal outreach morning of — planning to reach out to the 19 existing signups directly. not a blast, individual messages.

      comment velocity in first 4 hours — will be glued to PH all day tomorrow.

      no slack/discord upvote asks — wasn't planning on it but good to know PH actively detects this.

      launching 12:01am PT. going with conventional timing since most traffic so far has been US-based.

      and yeah — planning a behind-the-scenes update here regardless of outcome. the whole point of this post was transparency. would be weird to go quiet after.

  3. 1

    The pre-launch anxiety is real but those numbers look solid for a solo founder. One underrated PH tip: engage with every single comment in real-time during launch day. The algorithm rewards active discussion, and it signals to visitors that there is a real human behind the product. Also, consider posting a \"behind the scenes\" update on IH the day after — win or lose, the transparency builds trust.

  4. 1

    The "just ask ChatGPT" competitor framing is exactly right. Most validation tools get beat by "prompt engineering" in people's heads. The way to win there is to be so much faster and more structured that the convenience gap closes.

    On PH launch day: the comments matter more than upvotes for ranking. Genuine back-and-forth in your thread signals activity to PH's algorithm. Brief, specific responses from you to every comment in the first few hours makes a real difference.

    The 12% signup conversion from 84 visitors is decent. Most landing pages are 2-5% — you're in good shape if the traffic holds.

    1. 1

      the "convenience gap" framing is spot on. right now the workflow is: paste idea → wait 60 seconds → get verdict. chatgpt is faster for a casual "is this good?" but the structured output (market data, competitor analysis, kill criteria) is what you can't get from a chat.

      and good to know about comment velocity — will prioritize genuine back-and-forth over polished responses tomorrow. 12% conversion is encouraging but the real test is whether PH traffic converts the same way.

  5. 1

    This is a very short question but effective question. How fast do you give the 84 visitors you already have value? Yes, you want to grow but part of that is satisfying your current visitors through quick results and smooth onboarding to get them signing up and paying. The more effectively you do this, the more referrals from your users you can gain. Referrals are a form of growth too.

    1. 1

      you're right — i've been so focused on "get more traffic" that i haven't thought enough about the 84 who already showed up.

      the onboarding fix (killed email verification, oauth first) was step one. evaluation itself takes about 60 seconds which feels decent, but the pre-evaluation setup could be smoother.

      referrals as growth is interesting. might add a "share your verdict" feature post-launch — let people show off their GO or wear their KILL as a badge of honor.

      1. 1

        Happy to help! I also do funnel and onboarding audits for free, if you are interested then i can definitely assist you with that!

        1. 1

          appreciate the offer! will keep it in mind once i have more traffic to actually audit. right now the bottleneck is getting people to the funnel in the first place.

          1. 1

            No problem, looking forward to collaborating in future!

    2. 1

      This comment was deleted 2 months ago.

  6. 1

    The honest validation angle is what makes this interesting honestly.

    We build apps for founders and the most painful projects we worked on were ones where client had already convinced themselves the idea was great before talking to anyone real.

    They come with full requirements document, full enthusiasm, sometimes already spent money on designs. And somewhere in first week of development you can see the idea has fundamental problem nobody questioned.

    If someone had given them a KILL verdict early — would have saved months and a lot of money.

    The ChatGPT competitor concern is real but I think you have one advantage they dont. ChatGPT will find a way to say yes if you push it. Deterministic kill criteria that doesnt care about your feelings is actually the differentiator.

    Good luck with the launch today. Curious what your numbers look like by end of day 😄

    1. 1

      "full requirements document, full enthusiasm, sometimes already spent money on designs" — that's the scariest pattern. by that point even a friend won't tell you to stop.

      and yeah the chatgpt thing is the core bet. it's not that chatgpt can't analyze ideas. it's that it's designed to be helpful, not honest. push hard enough and it'll find a way to say "promising with some caveats."

      deterministic criteria that doesn't bend is exactly what i'm going for. thanks for articulating it better than i could.

      launching tomorrow — will share numbers here after.

  7. 1

    Two days out, the only numbers that have mattered for me were email signups from warm traffic and how many of those people actually came back the next day. PH upvotes feel nice, but they rarely predict revenue unless you already lined up your own audience first. If you have not done it yet, spend these last 48 hours on personal outreach, not launch page tweaks.

    1. 1

      solid advice. spending today on personal outreach to existing signups instead of tweaking the launch page. you're right that PH upvotes alone don't predict revenue — the real signal is whether existing users come back and tell someone else.

    2. 1

      This comment was deleted 2 months ago.

  8. 1

    This is really interesting.

    I’ve been thinking about this space while building something myself, and one thing I’ve noticed is that people struggle more with execution than the idea itself.

    I’m currently testing a small tool around this—not fully sure if it’s useful yet.

    Would you be open to taking a quick look and sharing honest feedback?

    1. 1

      sure — drop a link or DM me. always interested in what others are building around this.

  9. 1

    kill criteria is the right call. good luck thursday

    1. 1

      appreciate it. K1-K5 is the part i'm most confident about — everything else is still being debugged in real time. will report back with actual numbers.

    2. 1

      This comment was deleted 2 months ago.

  10. 1

    you named the real competitor yourself. "just ask chatgpt" kills most idea validators because people already trust the conversation format. the deterministic kill criteria (K1-K5) is the only defensible angle you have, but only if founders actually change their behavior after getting a KILL verdict. do you have any data on that? like how many users who got KILL actually stopped working on the idea vs just trying again with different inputs until they got GO?

    1. 1

      honest answer: i don't have that data yet. 9 signups, launched a week ago — too early to track behavior change.

      but the resubmission gaming concern is real. right now if someone tweaks the input to dodge a KILL, the same K1-K5 criteria still fire. the system checks for strong free competitors, market size, monetization path — not just how you phrase the idea. so gaming it is harder than it sounds.

      the deeper question — do founders actually stop after KILL — that's what i want to measure post-launch. if they don't, the tool is just expensive confirmation bias with extra steps.

      will share real numbers after thursday.

    2. 1

      This comment was deleted 2 months ago.

  11. 1

    This is super helpful — appreciate you sharing real numbers (not just “we launched and it went great”).

    One thing that always stands out to me with PH launches is how consistent the pattern is:

    Lots of traffic
    Solid signups
    But relatively low conversion to paid

    Curious — looking at your numbers so far, do you feel like:

    A) The traffic quality isn’t quite your ICP
    or
    B) The value prop just needs to be clearer on first visit?

    I’m prepping for a launch soon myself, and trying to think of PH more as:

    “top-of-funnel + credibility”

    rather than expecting it to drive revenue directly.

    Would love to hear what you’re seeing so far and what you’d do differently with 2–3 days left before launch.

    1. 1

      honestly probably B more than A. the landing page tries to do too much — i've been trimming it but "AI idea validator" still sounds generic until you see a KILL verdict in action.

      your framing of PH as top-of-funnel + credibility is exactly how i'm thinking about it now. not expecting revenue from launch day — expecting signal on what messaging actually lands.

      with 2 days left: i stopped adding features and started rewriting the first 3 seconds of the landing page. if that doesn't hook, nothing else matters.

  12. 1

    biggest thing I’d double check before launch: can someone understand what you do in 5 seconds? PH traffic is brutal — if it’s not instantly clear, people bounce fast

    1. 1

      good call — that's literally what i'm working on today. current tagline is "kill bad ideas before they kill your time" but i'm not sure that passes the 5-second test for someone who doesn't already know what idea validation means.

      testing a few variations before thursday. appreciate the nudge.

Trending on Indie Hackers
I've been building for months and made $0. Here's the honest psychological reason — and it's not what I expected. User Avatar 176 comments 7 years in agency, 200+ B2B campaigns, now building Outbound Glow User Avatar 53 comments This system tells you what’s working in your startup — every week User Avatar 52 comments 11 Weeks Ago I Had 0 Users. Now VIDI Has Reviewed $10M+ in Contracts - and I’m Opening a Small SAFE Round User Avatar 46 comments The "Book a Demo" Button Was Killing My Pipeline. Here's What I Replaced It With. User Avatar 23 comments My AI bill was bleeding me dry, so I built a "Smart Meter" for LLMs User Avatar 9 comments