Marketing a vibe-coded app

You built a working product. That is not the easy half being dismissed - it is genuinely the thing most people never manage. What you didn't get handed with it is that shipping and being found are two separate projects, and only one of them got faster.

The tools you used optimise for "does it run." Nothing in the prompt-to-app loop asks who this is for, what it replaces, or how anyone would ever encounter it. So that work never happened, and its absence is silent - no error, no failing test, no red text. Just a live URL with nothing on the other end of it.

It isn't a skills gap. It's an arithmetic problem, and it happens to be at its worst right now.

Supply exploded; attention didn't

One AI build platform alone reports on the order of a million new projects a week, and says roughly 63% of its users have never written code. The App Store added around 560,000 new apps in the first half of 2026. Meanwhile the number of results on a search page, slots in a feed, and names a model returns when asked "best tool for X" is exactly what it was before. The bottleneck moved from building to being found, and everyone crossed it at the same moment.

The default channels saturated on the same schedule

The places builders instinctively announce things - launch platforms, maker communities, the same three subreddits - absorbed that entire wave. They still work, but as one input among several, and they reward people who were already there.

Nothing you built is instrumented

The quiet killer. When a site is generated rather than assembled, analytics, Search Console and any definition of "a user did the thing" are absent unless you asked. So when the silence starts you can't tell whether nobody arrived, or people arrived and bounced, or people stayed and never paid - three different problems, three different fixes, and every decision from here is a coin flip.

Four questions, in this order. They take an afternoon between them, they cost nothing, and until all four are yes there is no point spending money or time on anything downstream.

A fifth, not quite a check: one obvious next action, repeated. Generated pages arrive with four buttons of equal weight, which is the same as having none.

Post-launch silence nearly always has one of five causes. They stack, so the order matters more than the effort you put into any one of them.

• Visibility - nobody arrives. Under roughly 100 visits a month you can't diagnose anything else; there isn't enough data for a conclusion to be real. The job is getting found at all, on one surface, for one thing you can genuinely own.
• Positioning - they arrive and bounce. Traffic exists and leaves in fifteen seconds. More traffic makes this worse: you're buying more people to confuse.
• Funnel - they get it and still don't convert. There's a specific step where people fall out, and it stays broken for months because nobody instrumented it.
• Measurement - you can't tell which of the above is true. The most common and most expensive gap, because it turns every decision into a guess. Fix it first when it applies; everything after gets cheaper.
• Loops - you convert but nothing compounds. Every customer costs what the last one did. Nothing about having users makes the next one easier to get.

Almost everyone works on gap three while their real problem is gap one. The free diagnostic exists to stop that specific mistake - five questions, result on screen, no email.

The single most useful thing anyone can give you here is a realistic curve, because the gap between what you expect and what happens is where most people quit.

Across four sites launched in 2026 and tracked in Search Console, impressions accumulated within days - between roughly 2,200 and 16,500 in each site's first weeks. Over the same period click-through rates sat between 0.1% and 0.6%, at average positions between 12 and 49. Read that honestly: the machines found and understood these sites quickly, and almost nobody clicked, because almost nobody clicks a result at position 22. Clicks come when position improves, and position improves with time. The impressions are the leading indicator; the clicks are the lagging one. All of it, including the unflattering numbers, is on the receipts page.

Rough timings: indexing takes days to a fortnight, real search positions take months, and AI citation moves faster than classic ranking because models weight relevance over accumulated authority. Talking directly to users answers you the same day and doesn't scale. You want one slow compounding channel and one fast answering one - the whole argument of the channel picker.

• Rebuild the landing page again. It feels like progress because it's the work you already know how to do. Without measurement, version four isn't better than version three - it's just newer, and you still can't tell.
• Treat a launch platform as the strategy. One day of heavily contested attention, no compounding. Worth doing, worth zero minutes of anxiety.
• Post your link into communities. Cold link-drops get removed, and removal is the good outcome - the bad one is being remembered as the person who did it. Be useful there for weeks first.
• Buy ads before the page converts. Paid traffic against a page nobody understands purchases proof you already had for free.
• Start five channels at once. Each gets a third of the effort needed for a signal, all five get abandoned around week three, and none of them teach you anything.

• If you're not sure anyone can find you at all: Is my site invisible to Google? - the 60-second tests, the causes, and the fixes ranked by effort. Do this first; it invalidates everything else.
• If you just shipped and don't know what today looks like: I built an app - now what? - an ordered plan for weeks one to four, including what to skip.
• If you want to know which gap is yours: the two-minute diagnostic, then the north-star metric picker, so you measure one honest number instead of fifteen flattering ones.
• If you want to know what normal looks like: the receipts - four real sites, dated Search Console numbers, published including the bad weeks.

One closing note, meant plainly: none of this is a judgement on how you built the thing. Shipping a working product in days with AI is a real advantage and it keeps getting better. The gap is that the same leverage doesn't exist yet on the distribution side, so the unglamorous half is still done by hand - by you, or by nobody.

The free diagnostic asks five questions and names the one gap causing your silence - visibility, positioning, funnel, measurement or loops - then gives you three moves in order. It takes two minutes and you see the result on screen.

Or get the 30-day post-build playbook Send it