How to do ASO: a weekly routine
ASO is not a one-time setup. It is a short loop: read your store reports every week, choose one change, ship it, and read what followed before the next change on that store. This guide walks through that loop for one app on both stores, with the checks that keep you from fooling yourself.
Store facts checked against 15 Apple and Google pages on · See the sources · How we check facts
In short
- Start from your own App Store Connect and Play Console reports, not from a keyword list.
- Find the step where people drop out: not shown in search, shown but not chosen, or viewed but not installed.
- Change one thing per store at a time, and write down the day it went live.
- Compare whole weeks before and after, from the same source. That is a before-and-after read, not a controlled test.
- When you need to know which of two versions does better, use the store's own randomized test.
Before week one: write down where you stand#
Spend an hour on a baseline. Without one, every later number floats.
- Your listing as it is today, in each language you publish: the title, subtitle and keyword field on the App Store; the title, short description and full description on Google Play; and the order of your screenshots on both.
- The last 28 days from App Store Connect: impressions, product page views and downloads, each split by source type. Write the dates next to the numbers.
- The same 28 days from Play Console: store listing visitors and install clicks, split by traffic source.
- Your average rating and how many new ratings arrived in that window, on each store.
- Anything else that moved downloads lately: a new version, a feature, a sale, a mention in the press. These are the first suspects when a number jumps.
Step 1: Read the search numbers#
Search is where a listing's words do their work, because they decide which searches can show the app. On the App Store, read impressions and downloads from the App Store search source.
On Google Play, generic search is harder to see. In Play Console's store listing reports, Google Play search counts searches for your app's name or a closely associated brand, and a search for a kind of app counts under Google Play explore, mixed with browsing. So the visitors your listing's words win from generic search sit inside Google Play explore. Read both sources, and keep them apart.
Read the week as a total, not day by day. A small app can see a few downloads one day and none the next, and that swing means nothing on its own.
Step 2: Find the step where people drop out#
Every listing is a funnel: the app is shown, someone opens the page or taps Get, and some of them install. Each step has its own fixes, so name the step before you name the change.
| What you see | Where to look | What to try |
|---|---|---|
| Few search impressions (App Store) or few store listing visitors (Google Play) | The words in the fields the store reads: title, subtitle and keyword field on the App Store; title and descriptions on Google Play | Use the words people type for what the app does. Drop words that repeat across fields |
| Seen in App Store search, rarely chosen | The search result itself: icon, title, subtitle and the first screenshots | Make the first screenshots say what the app does. See search impressions not becoming downloads |
| Many page views or visitors, few installs or install clicks | The rest of the page: screenshots, description, ratings, price and in-app purchases | Answer the question a visitor still has, and read recent reviews for a complaint that repeats |
| A word you use stopped showing the app | Whether the word is still in a field the store reads, and whether a rival changed its listing | See keyword in metadata not shown and a rival's listing changed |
The App Store's search step has a rate of its own, search result conversion: downloads from App Store search divided by the unique devices that saw the app there.
Where do candidate words come from? Neither store publishes how many people search for a word, so collect them from what people already say: your reviews, the stores' own search suggestions as you type, rivals' titles and subtitles, and, if you run Apple Ads, its search terms report for your campaigns.
Step 3: Choose one change#
Pick one change per store, aimed at the step that leaks most. Change two things at once and you cannot say which one the numbers followed.
Write it down before you ship it, in one line: what changes, why, which number you will read, and when. For example: "Replace the first screenshot with the timer screen, because few searchers choose the app; read App Store search conversion 14 days after it goes live."
NoteKeep the text and images you replace. If the new version reads worse, putting the old one back should take minutes, not an evening of digging.
Step 4: Ship it, and note the day it went live#
On the App Store, the app name, the subtitle and description and the screenshots are edited on an app version: they change when you submit a new version. Apple says App Review typically handles at least 50% of submissions in under 24 hours and 90% in under 48 hours.
Two shortcuts skip the new version. Promotional text can change without a new submission, though Apple says it doesn't affect search ranking. And when a product page optimization test finds better screenshots, you can apply that treatment to the product page; only an icon change waits for the next version.
On Google Play, you publish listing changes from Play Console. Google says certain apps may be subject to extended reviews of up to 7 days, or longer in exceptional cases.
The day that counts is the day the change appeared on the store, not the day you submitted or approved it. Open the public listing, check the change is there, and write that date down.
Step 5: Read what followed#
Give the store time to report the new days. Apple says data for a given day is considered complete two days after the reporting date, so read a window once its last day is a few days old. Then:
- Take whole weeks before and after the day the change went live, so both windows hold the same weekdays.
- Read the source the change was aimed at: App Store search for a keyword or screenshot change, not all downloads.
- Check the other sources. If browse traffic, or the other store, moved the same way, the season or a feature may be part of what you see.
- If you run Apple Ads, remember that App Store search includes the views and downloads from ads in search results, so a change in ad spend moves it too.
- Write the read down with the dates of both windows, even when it is unclear.
This is a before-and-after read, not a controlled test. It tells you what followed a change, not why. At a few downloads a day, only a large change stands out from the normal swing between weeks, so an unclear read is common, and saying so is honest.
When you need to know whether one version beats another, use the store's own randomized test: product page optimization on the App Store, which tests the icon, screenshots and app previews, and on iOS 27 and later the product page header, or store listing experiments on Google Play, which test graphics and, per language, the descriptions.
Step 6: Keep a change log#
A short log turns the routine into a record you can learn from. One row per change is enough:
- The store, the language and the field.
- The text or image before and after.
- The day it went live.
- What you read 7 and 14 days later, with the dates of both windows.
- Anything else that happened in those weeks.
After a few months, the log shows which kinds of change your own listing responds to. No general advice can tell you that.
Where RocketASO fits#
RocketASO runs this loop with you. It reads your App Store and Google Play reports, names the fix estimated to be worth most this week, and compares installs before and after it. It starts you with a draft of the copy or the keyword field, and nothing reaches the store until you approve it. To see where a listing stands first, run the free ASO audit.
Questions
How often should I change my listing?
One change per store at a time, left live long enough to read. RocketASO reads each change 7 and 14 days after it goes live, and a small app often needs the longer window before anything is clear.
Do I need a keyword tool to do ASO?
Not to start. Your store reports say how often the listing was shown and installed. A keyword tool helps you find words to try, but neither store publishes how many people search for a word, so any volume a tool shows is its own estimate.
Can I test two versions of my listing?
Yes, with each store's own tool: product page optimization on the App Store and store listing experiments on Google Play. Both split real visitors between versions at random, which a before-and-after read cannot do.
Sources
15 official pages, all checked . Each says what the page supports.
Apple (12)
- App Store Connect Analytics Help: Acquisition: The source types App Store search, App Store browse, app referrer and web referrer; App Store search includes views and downloads from search ads.
- Apple Ads Help: Glossary: The Search Terms tab reports search terms for an advertiser's own campaigns.
- App Store Connect Help: App information: The app name holds 2 to 30 characters and changes when you create a new version; the subtitle holds 30.
- Apple: Creating your product page: The subtitle and the description are updated when you submit a new version of the app.
- App Store Connect Help: Manage your App Store assets: Screenshots and previews are placed on a version page while that version is being prepared or was rejected.
- Apple: App Review: App Review typically reviews at least 50% of submissions in less than 24 hours and 90% in less than 48 hours.
- App Store Connect Help: Platform version information: Keywords hold up to 100 bytes; promotional text holds 170 characters and changes without a new submission; the description holds 4,000 characters.
- Apple: App Store search: Search weighs text relevance (title, subtitle, keywords, primary category) and user behavior such as downloads and ratings; promotional text doesn't affect search ranking.
- App Store Connect Help: Apply a test treatment to your product page: A treatment's screenshots and app previews can be applied to the product page; an icon change needs the next app version.
- App Store Connect Analytics Help: Download Analytics Reports: Data for a given day is considered complete two days after the reporting date.
- Apple: Product page optimization: A test compares up to three treatments of the icon, screenshots and app previews, for up to 90 days.
- Apple: Creative asset best practices: Creative assets appear in iOS 27 and iPadOS 27 and later, and product page optimization can test alternative header visuals.
Google (3)
- Play Console Help: Understand and grow your app's user base: Store listing reports count clicks since mid-2026; the definitions of visitors, install clicks, click-through rate, the traffic sources and the search term dimension.
- Play Console Help: Publish your app: Certain apps may be subject to extended reviews of up to 7 days, or longer in exceptional cases.
- Play Console Help: Run A/B tests on your store listing: Store listing experiments test up to 2 variants of graphics and, in localized experiments, descriptions; they stop after 6 months.