App store optimization: what Apple and Google document

App store optimization (ASO) is the work of making an app easier to find in App Store and Google Play search, and more likely to be installed once someone finds it. It has three parts: the text the stores read to decide what your app is about, the listing people see when they decide whether to install, and the quality signals the stores collect afterwards, such as ratings and stability. Neither Apple nor Google publishes how much each input weighs, so anyone quoting percentages is guessing.

A note on spelling. This article is written in British English, where the word is optimisation, but I use app store optimization for the term itself because that is how the stores and most searches spell it.

What is app store optimization?

It is the store equivalent of search engine optimisation, with some important differences. There are no links to earn. The text fields are tiny, so every word is a decision. How people behave after they see your listing counts. And there are two stores with different rules, so advice that is right for one can be wrong for the other.

In practice that means two questions, asked in order. Does the store understand which searches the app is relevant to? And when the app appears, do people choose it? Metadata answers the first. Screenshots, ratings and reputation answer the second, and the quality of the app feeds both.

The ranking inputs Apple and Google document

Both companies describe their search inputs in public, in general terms. Apple's App Store search page says results are based on text relevance, meaning matches for the app's title, subtitle, keywords and primary category, and on user behaviour, meaning downloads, ratings and reviews "and more". Google's guide to getting discovered on Google Play search says apps are ranked on a combination of ratings, reviews, downloads and other factors, and that the details are proprietary.

Input	App Store	Google Play
App name or title	Named by Apple as a text relevance input. Up to 30 characters.	Google advises a unique, focused title. 30 character limit.
Subtitle or short description	Subtitle is a text relevance input. Up to 30 characters.	Short description, 80 character limit.
Keyword field	Hidden field of 100 characters, separated by commas.	Does not exist.
Long description	Apple tells developers not to add keywords to it to improve search.	Full description of up to 4,000 characters. Google advises writing for users and warns against repeated keywords.
Ratings, reviews, downloads	Named as user behaviour inputs.	Named as ranking factors.
Stability	No published threshold tied to search.	Poor core vitals can reduce visibility.

Those limits are the ones in each store's documentation as I write this. Both stores have shortened limits before, so check the console before you plan copy around them.

App Store: name, subtitle and the keyword field

On iOS the text that counts is short: the app name, the subtitle and a hidden keyword field. Put the plainest description of what the app does into the name and subtitle, alongside the brand, because those are also what people read.

Apple's guidance for the keyword field is specific. Separate terms with commas and no spaces. Do not spend characters on plurals of words you already have, on generic words such as "app", on filler words, on competitors' names, or on trademarks you have no right to use. A common practice, which Apple does not spell out, is to avoid repeating words that are already in the name or subtitle so the field covers new ground.

Apple is equally clear that the long description is for people. Its guidance says not to add unnecessary keywords to the description in an attempt to improve search results. Apple also notes that in-app events and promoted in-app purchases can appear in search results, and that custom product pages can be assigned keywords so that a search shows the page that fits it best.

Google Play: title, short description and full description

Google Play has no keyword field. Google reads the listing itself, so the title, the short description and the full description do the work. Google's advice is to write for users, to say what they get from the app, and to put the important content near the top.

The limit on this is policy. Google states that repetitive or irrelevant use of keywords in the app name or descriptions can lead to an app being suspended. Write the description as clear prose that uses the phrases your users would use, once or twice, where they belong.

Ratings, reviews, retention and crash signals

Both stores name ratings, reviews and downloads as inputs. Google goes further in general terms, saying Play search factors in the overall experience of the app based on user behaviour and feedback. Neither store publishes a retention figure that feeds ranking, so treat claims about exact retention thresholds with suspicion. An app people keep is certainly reviewed better than one they delete.

On crashes, Google is explicit. Android vitals defines a set of core vitals, including the user-perceived crash rate and the rate of "app not responding" errors. Google publishes bad behaviour thresholds for them, and says that if an app exceeds one, Play may reduce its visibility and may show a warning on the store listing. Apple does not publish an equivalent threshold for search. On iOS, crashes reach your ranking through the one-star reviews they produce.

Screenshots, icon and conversion

Being shown is half of it. The other half is being chosen, and that is decided in a second or two by the icon, the name, the rating and the first images.

Apple states that, depending on orientation, the first one to three screenshots appear in search results when there is no app preview video. So on the App Store your first screenshots are effectively your advert, and most people will never see the fifth.

• One message per screenshot. Show the real interface doing one valuable thing, with a short caption that can be read at phone size.
• Lead with the outcome. The first image should answer "what does this do for me?". A logo or a splash screen does not.
• Keep the icon simple. It appears small, next to competitors, and has to be recognisable without text.

Both stores let you test this properly. App Store Connect offers product page optimisation, which shows alternative icons and screenshots to a share of visitors, and Play Console offers store listing experiments. Use them in preference to opinions in a meeting. Good listing artwork is a design job, and it belongs with whoever handles your product design.

Localisation

Both stores let you provide a separate listing for each language, and Google says directly that translations help search results and discoverability for users worldwide. Localisation is often the cheapest growth available to an app that already works, with three cautions.

• Research, do not translate. The phrase people search for in German or Arabic is rarely the literal translation of your English keyword. Have a native speaker choose the terms.
• Localise the screenshots too. Captions, sample content, currency and date formats should match the market.
• Do not promise a language the app does not speak. A localised listing in front of an English-only app earns installs, then uninstalls and poor reviews.

Start with the countries where you already see installs without trying. Both consoles report downloads by country.

What developers control in the build

ASO is usually handed to marketing, but several of the inputs are decided in the code. This is the part an app store optimization guide written by marketers tends to skip.

App size

A large download loses people on mobile data and people whose phones are nearly full. Engineers control this: ship Android App Bundles so each device downloads only what it needs, rely on the equivalent app thinning on iOS, compress images and media, download rarely used content on demand, and remove SDKs nobody uses any more. Both consoles report your download size, and it is worth reviewing at every release.

Crash rate and responsiveness

Given what Google says about core vitals, stability is a discovery issue on Android and a ratings issue everywhere. Crash reporting should be in the app before launch, not added after the first bad review. Test on older and cheaper Android devices, because that is where memory pressure and slow storage show up. Roll releases out gradually and watch the crash rate before going to everyone. This is ordinary observability and QA and testing discipline, applied to a phone.

Review prompts at the right moment

Both platforms provide an official in-app rating prompt, and both expect you to use it sparingly.

• On iOS, Apple says the prompt can be shown up to three times in a 365-day period, and recommends asking when users are most likely to feel satisfied, such as after completing a task, without interrupting them.
• On Android, Google enforces a quota it does not disclose, so a request may show nothing. Google therefore says not to tie the prompt to a button.
• Google also says the app should not ask questions before showing the card, including "Do you like the app?". The familiar pattern of sending happy users to the store and unhappy ones to a feedback form goes against that guidance.

The engineering task is to pick the trigger: after a successful booking, a completed workout, a delivered order. Never on first launch, never in the middle of a task, and never straight after an error.

Apple also lets you reset the summary rating when you release a new version. It recommends doing so sparingly, since having few ratings can put people off, and written reviews are not removed.

How to measure app store optimization

Use the stores' own numbers first. App Store Connect's analytics reports impressions, product page views, conversion rate and downloads, split by source such as search and browse. Play Console reports store listing visitors, acquisitions and conversion rate by traffic source, and can break search traffic down by search term.

1. Record a baseline. Impressions, conversion rate and installs from search, for each country that matters.
2. Change one thing at a time. If you change the title, the screenshots and the description together, you learn nothing.
3. Keep paid and organic apart. An advertising campaign changes both volume and conversion and will hide the effect of a listing change.
4. Give it time. Judge a change over weeks, across full weekly cycles, not after two days.
5. Measure what matters after the install. A keyword that brings installs which are deleted the same day is not a win.

Third-party ASO tools estimate keyword rankings and search volume. They are useful for direction, but the stores do not publish search volumes, so treat those figures as estimates.

Common app store optimization mistakes

• Stuffing the title. A name that is a string of keywords reads as spam to people and risks policy action on Google Play.
• Wasting the iOS keyword field. Spaces after commas, plurals, the word "app" and competitor names all burn characters Apple tells you not to spend.
• Using one store's tactics on the other. Keywording the App Store description does nothing Apple documents. Neglecting the Google Play description ignores the main text Google reads.
• Treating screenshots as a feature tour. Small text and six features per image convert worse than one clear claim.
• Asking for a rating at launch. You spend a limited prompt on someone with no opinion yet.
• Paying for reviews or installs. Both stores prohibit manipulating ratings, and the penalty can be removal.
• Ignoring stability. No listing change outruns an app that crashes.
• Doing it once. Competitors and store rules move. Review the listing with each significant release.

Frequently asked questions

What does ASO mean?

ASO stands for app store optimization: improving an app's visibility in App Store and Google Play search, and the share of people who install it after seeing the listing. It covers metadata, artwork, ratings and the technical quality of the app.

Is there an official app store optimization guide?

Yes, in effect. Apple publishes guidance on App Store search and on creating a product page, and Google publishes Play Console help on getting discovered in search and on Android vitals. Start with those before any third-party advice, because they are the only sources that speak for the stores.

Is an app store optimisation service worth paying for?

It can be, if the app is already stable and well rated and the service brings keyword research in your markets, strong creative and disciplined testing. Be cautious of any provider that guarantees a ranking or sells reviews. If the app crashes or has a low rating, spend the money on fixing the app first.

How long does ASO take to show results?

On the App Store, changes to the name, subtitle and keyword field go out with a new version, so the cycle is tied to your releases. On Google Play, listing text can be edited without a new build. In both cases I would judge a change over several weeks and not expect a single edit to transform installs.

Does ASO work without a good app?

Not for long. Ratings, reviews and, on Google Play, stability feed back into visibility, so a weak app loses the ground its listing gains. The order that works is a stable app, then a clear listing, then testing.

If the build side of this is where you are stuck, on size, stability or where the rating prompt goes, that is part of how we approach mobile app development, on both iOS and Android.