We measured our own App Store visibility. One app appeared, once.

The method, which costs nothing

Apple exposes a public search endpoint. You can query any term against a storefront and read back the ranked results, without a paid ASO tool and without a browser. We ran 27 commercial terms and 17 probes against the US storefront and recorded the full ranking for each.

Anyone can do this for their own app in an afternoon. If you have ever wondered whether your app appears for the term you think it targets, this answers it exactly rather than approximately.

Doing it for competitors at the same time is what makes it useful: you get the ranking, the title, and the rating count of every app above you.

The result: one hit out of twenty-seven

Of 27 terms, one returned an app of ours — `voice calorie counter`, at position 27. Everything else returned nothing of ours anywhere in the results. Combined with a website that had zero Search Console impressions at the time, that meant the studio had no working acquisition channel at all.

That is a bleak sentence to publish. It is also the only useful starting point, because the alternative is assuming the apps are findable and wondering why nothing installs.

It was not the ratings

Our first assumption was that zero ratings made ranking impossible. The data says otherwise: a competitor titled "AI Humaniser: Rewrite AI Text" ranks first for its own tokens with zero ratings, and the top ten for `humanize ai text` includes apps with three and six ratings.

So "we have no reviews, so ASO is hopeless" is measurably false, at least for retrievability. Ratings clearly affect whether someone taps your result. They were not what kept us out of the results entirely.

What did correlate was the title. Apps ranking for a term almost always had the term's exact tokens in their title.

The most expensive letter we have ever typed

Our humaniser app ranked second for `ai humaniser` — the British spelling, about 2,400 US searches a month. For `ai humanizer`, the American spelling at roughly 673,000, it did not appear in the first 98 results. Apple does not appear to stem across s and z, and every app in the top ten for the z-spelling uses that spelling.

Both titles are exactly thirty characters, so the swap would cost nothing in length. That is a factor of roughly 280 in addressable search volume, sitting behind one character.

We have not made the change yet — retitling is a product decision with brand consequences, not just an ASO one. But the gap does not shrink by being deferred; it is simply unrealised.

The mirror image: a category that is closed

Our receipt app is absent from 193 results for `ai receipt scanner` — a phrase that is a substring of its own name. The incumbents there hold 7.5 million, 1.35 million and 36,700 ratings. That category is not winnable by title work, and pretending otherwise would waste months.

So the two apps needed opposite channels: one is an App Store problem with an App Store fix, the other needed to be found through Google instead. That conclusion only became obvious once both were measured the same way.

And the studio name returns nothing

Searching `appvior` in the App Store returns no results. The seller name on our listings is a personal legal name rather than the studio's, so brand search does not work and neither does in-store cross-sell between our own apps.

For a studio publishing several apps, that is a compounding loss: every install is a one-off rather than a route to the next app.

What we would tell anyone else to do first

Measure before you optimise, using the free endpoint, and measure competitors in the same pass. Most ASO advice is generic because it is written without data; twenty minutes of querying replaces all of it with facts about your specific app.

Then check the boring things: does your title contain the words people actually type, in the spelling they type them, and is your seller name the brand you want searched. Those three cost nothing and, in our case, explained nearly everything.