Il nostro lancio su Product Hunt ha ricevuto 1 voto. Ecco il postmortem completo.

We launched FeedPulse on Product Hunt on 20 October 2026. It did not go well.

Most launch write-ups you read are from people whose launch worked. That makes them survivorship bias dressed up as advice. This one is the other kind. The numbers are unflattering and I am publishing them anyway, because a postmortem with the bad parts removed is just marketing.

The numbers

• Upvotes in the first 24 hours: 1. That was not a typo and it was not me.
• Comments: 0. The thread stayed empty through the entire reply window.
• Go-live: 08:01 UTC (12:01 AM PST), the standard slot.
• Topics: Developer Tools, SEO, Web Apps, Marketing.
• Preparation: roughly six weeks of maker-profile warm-up — hand-written comments on other launches, no automation.

For context on what was being launched: 18 embeddable widgets and 12 free SEO tools (10 checkers plus 2 generators). No paid tier, no signup, no cookies. So this was not a thin product shipped in a weekend. It still landed on Product Hunt with the impact of a pebble in a swimming pool.

What went wrong

1. We had no audience to launch to

This is the whole answer, really, and everything below is a footnote to it.

Product Hunt is a distribution amplifier, not a distribution source. It multiplies an existing audience. Multiplying zero gives you zero. We went in with no newsletter, no following, and a maker account whose entire history was six weeks of commenting on other people's launches.

The warm-up advice you read everywhere — comment genuinely for a few weeks before you launch — is not wrong, it is just radically insufficient. Commenting builds a little credibility with the handful of makers who read your comment. It does not build the few hundred people who will show up and upvote on command at midnight Pacific. Nothing builds that except having spent a year being useful somewhere people gather.

2. We launched on a Tuesday into a crowded board

Tuesday is one of the most contested days on Product Hunt. A product with no launch-day traffic of its own needs an uncontested slot far more than it needs a "good" one, because every position it gains is one it has to win on the leaderboard rather than bring with it. Launching into a quiet Saturday with 15 upvotes can out-rank launching into a Tuesday with 40.

3. The first comment was wrong, and we nearly shipped it worse

The prepared first comment claimed every public page held Lighthouse 100/100/100/100. That was false. The real split, measured: SEO 100, accessibility 100, desktop performance 98-100, best practices 96, mobile performance 85-88. A single commenter re-running PageSpeed Insights would have caught it in about thirty seconds, on a product whose entire pitch is honest SEO tooling.

We caught it and rewrote the comment before go-live to quote the real numbers and name the open problem — a React hydration mismatch that caps best practices at 96 and is the likely cause of a 3.7-4.1 second mobile LCP. That is still open today.

The lesson is not "check your claims". It is that launch copy gets written weeks early, in an optimistic mood, and then goes stale while the product changes underneath it. Every number in a launch asset needs re-measuring the week you launch, not the week you drafted it.

4. We spent launch-day attention on the wrong surface

Product Hunt outbound links carry rel="noreferrer noopener ugc". They are a citation and a referral source, not link equity. We knew that going in and still treated the PH page as the destination. The postmortem you are reading will almost certainly out-perform the launch itself on every metric we actually care about, because it can be linked to normally and it will still be here in two years.

What actually worked

The warm-up comments. Not for the launch, but on their own terms. Writing genuine answers on other makers' launches was the only part of the whole exercise that produced conversations with real people. If I had my six weeks back I would spend them the same way and simply not schedule a launch at the end of them.

Being forced to audit our own claims. Preparing to be looked at closely made us go through the product honestly, and we found things we would not otherwise have found:

• A fabricated aggregateRating of 4.9 from 127 reviews, hardcoded site-wide. There

was no reviews table, no 127 reviews, and the rating was not even displayed on the pages carrying it. Removed from all 1,319 pages. Fake review markup can cost you rich results across an entire site, and anyone can view-source and ask where the reviews are.

• A $9 Pro Lifetime price in a blog post for a tier that does not exist. It had

already been scraped into two directory listings by the time we found it.

• Our own deployment alias outranking our real domain in Google's top 10 for primary

keywords, because both served crawlable copies of the same content.

None of those were launch blockers. All of them were worse than the launch going badly.

What we would do differently

1. Earn distribution before trying to amplify it. A launch is the last step, not

the first. If you cannot name 200 people who would want to know, you do not have a launch, you have an announcement.

1. Re-measure every claim the week of launch. Not the week you drafted the copy.
2. Pick the quiet day. Rank is relative. Contested days punish products that arrive

without their own traffic.

1. Treat the postmortem as the real asset. It outlives the launch, it can be linked

to properly, and writing it honestly is the only part of this that compounds.

1. Do not cross-post the launch link. Stagger follow-ups by at least 48 hours and

lead with the lessons, never the "we're live" post. Communities punish the second thing and tolerate the first.

The uncomfortable conclusion

Product Hunt did not fail us. We arrived with nothing and received a proportionate result. The platform is a lever, and a lever applied to nothing moves nothing.

If you are a bootstrapped maker reading this while planning your own launch, the useful question is not "how do I run a good launch day". It is "who already cares, and how many of them are there". If the honest answer is nobody yet, spend the next six months fixing that and launch afterwards. The board will still be there.

We are taking our own advice. Our distribution strategy now is the slow kind: every site that installs one of our widgets carries an attribution link back, so the link profile grows with the product rather than with our marketing effort. That takes years rather than a day. After the launch we had, years feels like a reasonable trade.

---

FeedPulse is 18 free embeddable widgets and 12 free SEO tools — no signup, no cookies, no paid tier. If you want the hit counter or the live traffic feed, they are free and always will be.