FR

Get early access

App Store · Google Play · Rankings

How to check an app's keyword ranking, and get a number you can trust

By the ASO Gauge editorial team Published Facts checked on 26 October 2026 9 min read

To check an app’s rank for a keyword, search the keyword in the store, in the country you care about, on the kind of device your customers use, without your own history, and count down to your app, skipping the ads. Then search again: ranks move within a day and from one day to the next. A rank you can trust is the median of several searches made the same day, kept with its spread, read to the same depth every time (the App Store returned up to 250 apps per search, Google Play’s web search up to 30), and written down as absent or corrected when that is what happened.

Every step below comes with real searches made on 25 and 26 October 2026 for a habit and routine app, the one on our home page.

What is a keyword ranking, exactly?

A ranking is your app’s position in the ordered list a store returns for one search. That list belongs to one store, one country, one language, one kind of device and one moment. Change any of them and the list can change, so a rank given without them is half a fact.

Apple lists what goes into the order:

The parameters used for app ranking and discoverability are customer engagement on the App Store, such as the number and quality of ratings and reviews, application downloads and date of launch on the App Store, the relevance of text and titles, keywords and descriptive categories provided in the app metadata.

Documented · Apple Learn more about search results and ranking

Google adds that the weights depend on where the search happens:

These main factors impacting ranking are weighted differently based on where on Google Play a user is looking, the device they are on, and their personal preferences.

Documented · Google App Discovery and Ranking

Why do App Store and Google Play rankings differ from one person to another?

Five things change the list. Four of them were measured for this guide.

  • The country. Each country has its own store and its own results. Apple’s ratings, one of its ranking parameters, are kept per country: the summary rating “is specific to each territory on the App Store”. On 25 October, the same app was 5th for “habit tracker” in the United States, 11th in the United Kingdom, 21st in Germany and 7th in France.
  • The language. A Google Play search carries a language as well as a country. In France on 26 October, a search for “routine” returned 27 of the same 30 apps with English and with French as the language, only 7 of them at the same place. On the App Store, a country searches the languages Apple lists for it.
  • The device. Google names “the device they are on”, and the App Store’s iPhone and iPad results differed when we compared them (below).
  • The person. Apple writes that “your searches, browsing, purchases, and device trust score may be used to personalize your experience on the App Store”, and Google cites “personal preferences”. Neither says how much a search changes from one person to the next.
  • The moment. Apple says it is “constantly evolving how search works on the App Store”, and Google that it constantly tests and optimizes discovery. Ranks also move within a day, as shown below.

How do you check a ranking by hand?

  1. Choose the country. The App Store on your phone follows your Apple Account’s country or region, and Apple asks you to spend your balance and cancel some subscriptions before changing it. For another country, Apple’s web App Store shows the first results without an account, at an address such as apps.apple.com/fr/iphone/search?term=routine.
  2. Choose the device. Read iPhone results for an iPhone audience, iPad results for iPad.
  3. Leave your history out. Apple lets you turn off Personalized Recommendations (Settings, your name, Media & Purchases, View Account). Google lets you turn off Web & App Activity. A browser that is not signed in carries no account.
  4. Skip the ads. Apple says its ads carry a prominent “Ad” mark, and Google labels them “ads” or “sponsored”. Count only the other results.
  5. Write down four things: the time, the rank, how many results the store returned, and whether it searched another word than yours.
  6. Search again, twice, later the same day. The next section shows why.

So the web page is a fair way to read the top of another country’s results by hand. It stops there.

Why is one search not a rank?

Because the order moves while nothing visible changes.

A single search at 20:45 would have reported 10. The evening’s median was 7. After midnight, the median was 10. Nothing in these readings says why the order moved, and that is the point: with one search, you cannot tell a move from a wobble. Some keywords hold still, others do not, and neither store documents which.

How do you turn several searches into one rank?

Keep two numbers per keyword, per country, per day.

  • The median, the middle of three searches. One odd search does not move it: the median of 7, 10 and 7 is 7.
  • The spread, the best and the worst rank of the day. A spread of 7–10 says the rank is not sharp that day; 5–5 says it is.

Why not the average? The average of 7, 10 and 7 is 8, a rank the store never showed. The median is always a rank that was really seen.

Space the searches through the day, and keep the same number every day: a day with one search and a day with three are not read the same way. Then compare days by their medians, and judge any move against the spread. The guide to judging a listing change turns this into a decision rule.

How deep should you read the results?

To the same depth every time, as deep as the store goes.

  • App Store. The App Store app’s search returned up to 250 apps per search: between 234 and 250 for the keywords measured on 25 October. Below that, an app is not in the results.
  • Google Play. Google Play’s web search returned up to 30 apps. The list length moves too: for “habit tracker” in the United States, it returned 21 apps at 00:09 on 26 October, then 30 at 00:13 and 00:18.

A day read deeper than the others looks like a gain that never happened. Two shortcuts read less than they seem:

  • Apple’s web App Store matched the app’s order, but shows 11 or 12 apps per search.
  • The iTunes Search API is a public Apple interface, sometimes used to check ranks. Apple describes it otherwise: “The Search API allows you to place search fields in your website to search for content within the iTunes Store and Apple Books Store.”

Read ranks from the App Store app’s own search, not from that API.

What if your app is absent, or the store changed the word?

Some searches end without a rank. Write down what happened instead of a number.

  • Absent. The app is not in the results the store returned. Write “not in the first 250” (or 30 on Google Play), never 251, and leave the day out of any average. On 25 October, the app was absent for “hábitos” in Brazil in all three searches, out of 249 results.
  • Corrected. The store searched another word and said so. Your rank then belongs to that other word: keep it apart. The same evening, “morning rutine” came back as a search for “morning routine” in all three searches, with the app 1st in those results. The spelling correction guide explains when this happens.
  • Failed. No answer, a blocked request, a changed page. Write the reason, and do not fill the gap with yesterday’s rank.

A real evening of readings

Here is what this method gives, for the same app on 25 October, from three searches per keyword between 20:39 and 20:52 Paris time.

SearchStore and countryThree searchesThe day’s reading
habit trackerApp Store, United States5, 5, 55
habit trackerApp Store, France7, 10, 77, spread 7–10
habitsApp Store, United Kingdom79, 79, 7979
hábitosApp Store, BrazilAbsent three timesNot in the 249 results
morning rutineApp Store, United StatesCorrected three timesSearched as “morning routine”, 1st there
tagesroutineGoogle Play, Germany29, 28, 2929, spread 28–29
habit trackerGoogle Play, United States14, 14, 1414, out of 21 results

Each line says what happened, including when there is no rank.

How do you track rankings over time?

The same way, every day: same countries, same device, same depth, same number of searches. Done by hand, 20 keywords in 5 countries with three searches each is 300 searches a day.

That routine is what ASO Gauge runs: three real searches a day per keyword and country, the median and its spread, 250 App Store results and 30 Google Play results read every time, and one of six states for every cell, never a blank. Its method is on our home page.

Whatever you use, ask three questions of it: how many searches make one rank, how deep it reads, and what it does when the app is absent or the word is corrected. The answers decide whether its numbers mean anything.

Common mistakes

  • Checking on your own phone, signed in, once. That is one sample, shaped by your history.
  • Mixing iPhone and iPad. Their lists differ.
  • Using the iTunes Search API for ranks. Its order is not the App Store’s.
  • Writing 251 for an absent app. It is not a rank, and it bends every average.
  • Comparing days read at different depths. A deeper day looks like a gain.
  • Ignoring a correction. A rank for the corrected word is not a rank for yours.
  • Quoting one number for all countries. Each storefront has its own rank.

Sources

Checked on 26 October 2026.