How to find long-tail keywords you can actually rank for

“Go for the long tail” is ASO's universal fallback advice. It gets handed out without ever being checked, as if lengthening a phrase were enough to make competition evaporate. We put it to the test.

On 23 August 2026 we took 6 heavily contested keywords on the French Play Store, asked Play Store autocomplete for its own variants — real user queries, not invented combinations — and measured the entry ticket for each: the review count of the least established app in its top 10. 24 variants in all.

The advice holds two times out of three

And when they are, the gain is more modest than advertised: the median factor is 2.5. Not ten, not a hundred. Lengthening a phrase typically divides the entry ticket by two or three — useful, but it does not turn an unreachable keyword into an easy target.

The case that should serve as a warning: “prise de notes” has an entry ticket of 6,471 reviews. Its variant “prise de notes gratuit” climbs to 44,427 — seven times harder than the phrase it derives from. Lengthening had exactly the opposite effect here.

Qualify the use, not the price

Splitting the variants by what they add makes the pattern clear — and it condemns the most common reflex.

Adding “free” changes nothing: zero median gain. Which makes sense once you think about it — the apps dominating “photo editing” are already free, so they also dominate “free photo editing”. The word excludes nobody.

A qualifier that narrows the use, on the other hand, divides the entry ticket by five at the median. “note taking” → “voice note taking”: from 6,471 to 1,205 reviews. “photo editing” → “AI photo editing”: from 429,548 to 13,629. “meditation” → “guided meditation”: from 20,419 to 4,029. Every word you add should exclude competitors, not describe a business model everyone shares.

The gain is proportional to how hard the head keyword is

This is the most useful finding of the study, and the least intuitive: the long tail does not pay the same everywhere. It pays where the head keyword is overwhelming, and almost nothing where it is already affordable.

“vpn” demands 904,770 reviews to enter its top 10; its best variant demands 13,988, sixty-five times less. At the other end, “budget” only requires 632 reviews, and none of its variants does better: there was nothing to gain, because there was nothing to escape.

The practical consequence is clear: before hunting for a variant, measure the head keyword. If it is already within reach, lengthening the phrase will cost you search volume without buying any accessibility. The long tail is a remedy, not a routine.

The method, in three steps

First, measure the entry ticket of your head keyword — the least established app in its top 10. If it is within reach, stop there: you have your target.

Otherwise, type your keyword into the Play Store search bar and read the suggestions: those are real queries, unlike phrases invented at your desk. Discard the ones that only add a price (“free”), keep the ones that narrow the use, the format or the audience.

Finally, measure the entry ticket of each candidate — one in three will be harder than your starting keyword, so check rather than assume. Keep the one with the lowest ticket among those that honestly describe your app: ranking on a query that does not match what you do earns you installs that uninstall.

The reference thresholds are here: how many reviews to enter a top 10, and the full difficulty calculation is published here.

Method and limits

6 head keywords, 24 variants, French Play Store, 23 August 2026. Variants come from Play Store autocomplete and all contain the head keyword. Listings whose review count Google does not publish are excluded from the calculation, never counted as zero.

The sample is small: twenty-four variants across six universes. The gaps between groups (“free” versus usage qualifier) are pronounced enough to report, but with only five variants in the first group, treat the number as a strong lead rather than a law. We will widen it.