I built it in a weekend. How does anyone find it?

You shipped. That is the part most people never do.

Now the honest problem. About 1.3 million new iOS and Play apps came out in the first half of 2026, double the year before. When someone asks ChatGPT for an app like yours, it names the ones it has read about.

Your app has a store listing, a landing page and nothing else for the AI to read. We fix the “nothing else”. You keep building.

The odds for a new app

The numbers are rough. Lovable surveyed more than 14,000 users. 10.7 percent earn income from what they built. More than 60 percent want to and have not yet.

RevenueCat counted more than 14,700 new subscription apps a month in early 2026. Apps launched in 2025 or later earned about 3 percent of subscription revenue.

Most first customers still come from people you know: 56 percent, in Supabase’s 2026 survey of 2,000 builders. Launch sites gave 6 percent.

The growing channel is a stranger asking an assistant “what app should I use for X”. The assistant gives a shortlist of three. It builds that list from reviews, Reddit threads, YouTube videos and roundup articles. New apps are on none of them by default.

The questions we would ask for you

Real prompts are short. Two-thirds are under 15 words, and a third include something personal about the asker. For an app, three of the five free-check questions usually look like this:

1. best app for [the habit or job], for example “best app to track freelance invoices on iPhone”
2. free [app type] app with no ads, or the price-shaped version for your category
3. I have [a situation], what [app type] actually helps, the persona prompt, where the AI reads reviews and threads most closely

Each one runs three times on ChatGPT, Gemini, Claude, Perplexity and Muse: 75 answers. For every answer: were you named, who was named instead, which pages were cited.

Buyers who are other builders? We weight Claude, since 59 percent of startup builders pay for it. Buyers who are regular people? ChatGPT and Gemini matter most.

What an app project includes

AI reads different sources for apps than for B2B software: store reviews, Product Hunt, Reddit, YouTube demos and “best apps for X” roundups. Ahrefs measured 75,000 brands. YouTube mentions had the strongest correlation with AI visibility of anything it measured.

So a project for an app is built around those sources:

• 25 named lists and sites pitched, two follow-ups each. These are the roundups and “best apps” articles that actually appeared in your citations.
• Three pages written for you. Your story page: what it does, who it is for, what it costs, who built it and why. One honest comparison with the app the AI keeps naming. One alternatives page. These answer the AI’s own searches.
• Review profiles on the two sites AI cites most for your category. Often Product Hunt and the App Store or G2. Claimed, accurate, with a plain plan for asking real users for honest reviews. In a 2026 experiment, a rating edge of under 0.1 stars was enough to beat a better-known brand.
• A founder-led Reddit plan: the threads that were cited and a useful disclosed answer for each. Plus each subreddit’s rules, so you are not banned on day one.
• A trust checklist. A scan of 5,600 public vibe-coded apps found more than 2,000 high-impact vulnerabilities, and buyers have started asking. If you want, one of the three pages is a clear privacy and security page. It is not a security audit. It is you saying plainly what you store and what you do not.
• Re-checks at day 30, 60 and 90 on the same questions.

Proof, and what we cannot show yet

We are new. Today we can show you the full method and a real sample report, built the way yours would be. It has the prompt-level detail: who got named instead, and the three to five sources behind them. The research behind every number is linked at the bottom of each page.

We cannot show a client before and after yet. The founding cohort of 20 gets the audit free with the project. In exchange, we may publish an anonymised before and after, misses included. They will appear here as they land.

We would rather show you an honest 4 of 60 to 11 of 60 than a screenshot.

What we will not do

• Guarantee mentions or rankings, ever.
• Write fake reviews, run persona accounts or hide text. Reddit only under your own name.
• Blast directories. ChatGPT and Gemini rarely cite them.
• Hide paid placements.
• Put you on a retainer by default. One price, once.

Price

Every deliverable ships within 90 days or your money back.

The day-60 target: named in at least your baseline plus 3 of the 20 agreed answers. Never fewer than 3, never more than 7. Miss it and we keep working at no charge until day 120. We never promise what an AI will say.

No checkout at launch. You fill in a short request form, I confirm fit within 2 working days, then send an invoice.

You built it. Now the machines need to know it exists.

Questions