How many reviews does your app need to rank on Google Play?

It is the first question a developer asks when looking at a keyword: “do I have enough reviews to stand a chance here?” Search for the answer and all you will find is “it depends.” True, but useless.

So we measured it. On 23 August 2026 we pulled the top 10 for 30 keywords across 6 categories of the French Play Store, and counted the reviews of every ranked app. Here is what the numbers say — including where they contradict intuition.

The single number, and why it is almost useless

That is the short answer: about 7,012 reviews. Keep it, but it is close to unusable on its own — because the spread between keywords is enormous.

On “conjugaison”, the least known app in the top 10 has 10 reviews. Ten. On “vpn”, the least known has 904,757. Between two keywords, the entry ticket varies by a factor close to 90,000. Any general answer to “how many reviews?” is therefore wrong for almost every actual case.

The good news is that the question is badly framed. It does have a useful answer — but keyword by keyword, and not with the number people usually look at.

The number that matters is the minimum, not the average

Across our 30 keywords, the median app in the top 10 has 133,490 reviews. The median entry ticket is 7,012 reviews. That is a factor of 19 between the two.

That gap explains why so many developers give up for the wrong reason. Looking at “the average level of the top 10” paints a crushing picture — hundreds of thousands of reviews — when the real question is entirely different: is there, in this ranking, at least one app as unestablished as mine?

If yes, you have an existence proof: on that specific keyword, the algorithm accepts an app of your size. If no — if even the weakest app in the top 10 exceeds you by an order of magnitude — no listing optimization will close that gap in the short term.

How it breaks down by category

Five keywords per category. The entry ticket is the median of the minimums; the right column shows the median level of the top 10, to gauge the gap.

The gap between categories is brutal: a median of 56 reviews is enough to appear in an Education top 10, against 291,774 in Photography. This is not about quality — photography has attracted apps with tens of millions of installs for a decade, and French verb conjugation has not.

One important caveat: a category median describes no single keyword. In Photography, the entry ticket ranges from 108 reviews to over 429,000 depending on the keyword. Even in the most locked-down categories, open doors remain — you just have to look for them one by one.

One keyword in three is open to a small app

This is the most encouraging finding: of our 30 keywords, 10 have an entry ticket below 1,000 reviews, and 4 below 100. In other words, in a third of cases a modest app already sits in the top 10 — living proof that it can be done.

The reverse is also true: 14 keywords out of 30 have an entry ticket above 10,000 reviews. For a young app those are not targets, they are walls. Aiming at them anyway is the most common way to lose six months.

How to apply this to your app

The method takes three steps and needs no tool: open the Play Store in your target country, type your keyword, then look at the first ten results and find the one with the FEWEST reviews. Compare that number to yours.

If it is in the same range as yours, or lower, the keyword is playable: build your title and description around it. If it exceeds you by a factor of ten or more, move to a longer, more specific keyword — that is where the two-digit entry tickets live.

What this study does not say

Thirty keywords across six categories is a sample, not a census: each category median rests on five keywords and should be read as an order of magnitude. The measurements date from 23 August 2026, and rankings move.

Above all, review count is not the only ranking factor. Keyword presence in the title, density in the description, download volume, retention and Android vitals (crashes, ANRs) all weigh in. Review count has one particular merit: it is the only one of those signals Google publishes, so the only one you can measure on your competitors without guessing.

One last methodological note: Google does not publish review counts for some listings, even heavily installed ones. Those were excluded from the calculations rather than counted as zero — counting them as zero would mechanically collapse the entry tickets and make every keyword look falsely accessible.