App Store Keywords: How to Research, Score and Place Them on iOS and Google Play
Six steps, one scoring template and the per-store placement rules with their official character limits. Run first-hand on a published app, with the rank readings to show for it.
App store keywords decide whether your app appears when someone searches, and the two stores treat them differently. iOS reads your name, subtitle, a hidden 100-character field and your category. Google Play has no keyword field at all; it reads your listing text. The workflow is the same on both: seed, expand, score, place, localize, track. This guide runs it on a real app and gives you the template, so the next run is yours.
What This Guide Covers
- Per-store placement rules with character limits from Apple's and Google's own pages
- Where candidate keywords come from, with our own captures of both stores' suggestions
- A scoring template you can copy into a sheet, filled in for a real app
- A tracking table with rank readings from Apple's search API, dated
- What to do with the same keyword list off the stores: landing page and ads
Every character limit below is quoted from Apple or Google, linked and dated. Apple does not publish search volumes, so where the guide uses a volume signal it says what the signal is. The first-hand run uses a customer's published apps and only public data; no listing was changed. The wider ASO picture (icons, screenshots, ratings, custom product pages) is in the ASO guide; this page is the keyword workflow only.
TL;DR Quick Summary
App store keywords are the words the stores match a search against. On iOS that is your 30-character name, 30-character subtitle, a hidden 100-character keyword field and your category, per Apple. Google Play has no keyword field and indexes your listing text: a 30-character name, an 80-character short description and a 4,000-character description. The workflow: seed ten phrases from what the app does, expand them with each store's auto-suggest and competitor listings to about 100 candidates, score each on relevance, volume signal and difficulty and keep the top 20, place them to the character with no word repeated across iOS fields, localize a separate set per storefront, and track ranks weekly for a month, then monthly. The template and a real run are below.
Publish your app free →Table of Contents
Six steps in order, the placement rules first because they decide what the steps are for.
- How each store reads keywords: the placement rules
- Step 1: Seed
- Step 2: Expand
- Step 3: Score
- Step 4: Place
- Step 5: Localize
- Step 6: Track
- First-hand: the workflow run on a published app
- Mobile app keywords beyond the stores
- What are the most popular keywords on the App Store?
- Six keyword mistakes that cost rankings
How each store reads keywords: the placement rules
Everything downstream depends on where each store looks, so here it is in one table. The iOS rows come from Apple's search and product page documentation; the Google Play rows from the store listing help pages. Where a store does not document something, the table says so rather than repeating what tool vendors assume.
| Field | iOS (App Store) | Google Play |
|---|---|---|
| App name | 30 characters. Apple: "Your app's name plays a critical role in how users discover it." | 30 characters. Indexed. |
| Subtitle | 30 characters, indexed. Appears under the name everywhere in the store. | No subtitle field. |
| Keyword field | 100 characters, comma-separated, no spaces after commas. Hidden from users. Apple: do not repeat words already in the name, subtitle or category; plurals count as duplicates; no generic words like "app"; no competitor app names or trademarks. | None. Google Play has no keyword field. |
| Short description | No equivalent. Promotional text (170 characters) is explicitly not used for search ranking. | 80 characters. Google: it "can appear in locations beyond your store listing". Treated as indexed by every ASO tool; Google does not publish the weighting. |
| Description | Not stated as indexed. Write it for humans. | 4,000 characters. Google warns that repetitive or irrelevant keywords "will not impact ranking" and can get a listing suspended. |
| Category | Indexed, per Apple (primary and optional secondary). | Indexed. |
| Ratings | "Can influence how your app ranks in App Store search", per Apple. | A ranking signal; not documented in detail. |
Sources: Apple, App Store search and product page guidance; Google, store listing character limits, metadata policy and store listing best practices. All read 24 September 2026.
The hidden iOS keyword field gets one paragraph here because the ASO guide covers it in full. The rules that matter for this workflow: 100 characters total; separate terms with commas and no spaces; use single words and let Apple combine them ("board,exam" matches "board exam"); never repeat a word that is already in your name, subtitle or category, because Apple already indexes those and the characters are wasted; and never put a competitor's app name in it, which is grounds for rejection.
Step 1: Seed: write down what the app is for
A seed is a phrase a buyer might type before they know your app exists. Ten is enough to start. Take them from four places, and write down which place each came from, because that column becomes your relevance score in step 3.
- The job: what the app does, in the buyer's words. For a study app: "board exam practice", "cpa review", "accounting quiz".
- The category: the noun the store uses. "Education" is indexed on iOS, so you do not need it as a keyword; the phrases inside the category ("exam prep", "flashcards") you do.
- The reviews: your own and your competitors'. People describe the app in the words they searched with. "mock exam" and "practice questions" came from reviews in our run.
- The competitors: the names people compare you with. Write them down for step 2 and then take them out; on iOS a competitor's app name in your keywords is grounds for rejection, and on Google Play unattributed use of another brand breaks the metadata policy.
One rule for the seeds: two words minimum where you can. In step 2 you will see why: a single-word seed like "cpa" returns "cpap" and "cpanel" from both stores, which tells you nothing.
Step 2: Expand: let the stores tell you what people type
Now let the stores expand the list. Both have an auto-suggest, and both suggest from what people actually search, which makes them the only free volume signal you have. Type each seed into the App Store search box on an iPhone and into Google Play, and write down every suggestion. Twenty minutes gets you from ten seeds to a hundred candidates.
The App Store's suggestions for our seeds, read the same day: for cpa exam it offered "cpa exam prep 2026", "becker cpa exam review", "cpa exam prep – mi. accounting", "cpa exam ar" and "cpa exam ai"; for cpa review, "cpa reviewer" and "applied accountancy cpa review"; for accounting exam, "uworld accounting – exam prep", "cma accounting exam prep" and "financial accounting exam"; for cpale, the exam our app is built for, "cpale reviewer" and "cpale reviewer 2026". Three things to notice, because they are the same on every run:
- Suggestions are ordered. The first suggestion for a seed is the most searched completion. That order is your volume signal in step 3: position 1 to 3 scores 3, 4 to 6 scores 2, anything lower scores 1.
- Half the suggestions are app names. "becker cpa exam review" and "uworld accounting" are competitors. They tell you who owns the phrase; they are not keywords you may use.
- The two stores differ. Play offered "cpa exam prep free"; the App Store did not. Keep the two lists separate; you will place them separately in step 4.
Then mine the competitors themselves. Open the top ten listings for your best seed and copy their name, subtitle and first two lines of description into the sheet. Phrases that repeat across three or more listings are phrases they have already tested. The free tiers of AppTweak, AppFollow and Sensor Tower will do this scraping for you and add a volume estimate; treat the estimate as a signal, not a count, because Apple does not publish search volumes and every vendor models them. The ASO guide compares those tools; this page does not rank them.
Step 3: Score: cut the list with a template
A hundred candidates is too many to place, so score them. The template below has five columns you fill and one it computes. It is deliberately coarse: three-point scales, no decimals, done in half an hour. The point is to force a decision per phrase, not to model the store.
| Column | How to score it |
|---|---|
| Relevance (1 to 3) | 3 = the phrase describes what the app does for the person searching. 2 = adjacent (they would be happy to find you). 1 = a stretch. 0 = wrong app, or a competitor's name. |
| Volume signal (1 to 3) | From step 2: suggest position 1 to 3 scores 3; 4 to 6 scores 2; lower, or only in a tool estimate, scores 1. |
| Difficulty (1 to 3) | Open the store search for the phrase. If the top three results each have thousands of ratings, score 3; hundreds, 2; fewer, 1. Ratings count is the one competitive signal both stores show you. |
| Current rank | Where you sit today for the phrase, if anywhere. Blank on a new app. Read it the way step 6 describes. |
| Score | Relevance × 2 + volume + (4 − difficulty). Maximum 12. Keep at 10 or more, maybe at 8 to 9, drop at 7 or less. |
Filled in for the app in our run, from the candidates step 2 produced:
| Keyword | Relevance | Volume | Difficulty | Score | Call |
|---|---|---|---|---|---|
| cpale reviewer | 3 | 2 | 1 | 11 | Keep |
| cpa review | 3 | 3 | 3 | 10 | Keep |
| board exam | 3 | 2 | 2 | 10 | Keep |
| cpa practice questions | 3 | 1 | 2 | 9 | Maybe |
| cpa exam prep | 2 | 3 | 3 | 8 | Maybe |
| accounting exam | 2 | 2 | 2 | 8 | Maybe |
| cpa mobile | 1 | 2 | 1 | 7 | Drop |
| cpap | 0 | 3 | 1 | 6 | Drop |
| becker cpa exam review | 0 | 3 | 3 | 6 | Drop: competitor name |
Two things the scored table shows that the raw list did not. "cpa exam prep" has the best volume signal on the sheet and still lands in Maybe, because the app is built for the Philippine CPALE and only partly serves a US CPA candidate: relevance 2 halves its score. And "cpale reviewer", a phrase with a weaker volume signal, scores highest, because nobody with real ratings owns it. That inversion, volume down and relevance and ease up, is what the template is for.
Step 4: Place: fill each store’s fields to the character
Placement is where the two stores part ways. On iOS you have three indexed slots with hard character limits and a rule that no word may appear twice across them. On Google Play you have three text fields that people read, and the words have to earn their place in a sentence. Fill iOS first, because its limits are tighter and they force the priority order.
iOS: name, subtitle, keyword field
- Name (30): your brand plus the one phrase with the highest score that a buyer would recognise. Apple front-loads the name in results, so the phrase goes after the brand, not before it, unless the brand is unknown.
- Subtitle (30): the next two keepers, written as a phrase a person would read under the name. No word from the name repeated.
- Keyword field (100): every remaining keeper as single words, comma-separated, no spaces, no plurals, no word already in the name, subtitle or category. Fill all 100 characters; Apple does not reward the unused ones.
- Do not: put a competitor's name anywhere, use "app" or "free" as keywords (Apple calls them too generic and matches "app" by default), or stuff the name: "CPA Reviewer Exam Prep Quiz Practice" is rejected as much as it is ignored.
Google Play: name, short description, description
- Name (30): the same rule as iOS. The two stores read your name separately, so it may differ, but a different name across stores costs you brand searches from people who saw the other one.
- Short description (80): your best two or three keepers in one sentence a person would act on. This line shows in search results and Google says it can appear beyond your listing, so it is doing conversion work as well as indexing work.
- Description (4,000): the keepers used naturally, each a few times, the most important ones in the first 250 characters. Google's own guidance: unnecessary keywords "will not impact ranking" and repetition can get the listing suspended. Write for the reader; the index follows.
- Do not: repeat a phrase every paragraph, list keywords as a block at the end, or use a brand you do not own. All three are named in the metadata policy.
The fields live in App Store Connect and the Play Console; the App Store submission guide walks through where each one is. If you publish through Appy Pie AI, the same fields are collected on the publish form and submitted for you, but the words are still yours to choose, and this is the step where you choose them.
Step 5: Localize: one keyword set per storefront
Both stores let you localize the listing per storefront language, and on both the localized fields are indexed for searches in that language. That makes localization a keyword step, not a translation step: a French buyer does not type the French for "budget tracker"; they type what French apps in the category are called. Run steps 1 to 4 again per language, with the auto-suggest switched to that store, rather than translating the English set.
- Where it lives: on iOS each localization has its own name, subtitle and keyword field; on Google Play each language has its own name, short and full description. Translate the description for the reader; research the keywords for the searcher.
- Which languages first: the ones your reviews and analytics already show. A study app for a Philippine exam does not need Spanish; it may need Filipino search terms alongside English on the same storefront.
- The cross-localization trick on iOS: ASO vendors report that some storefronts index more than one localization (the US storefront reading Spanish (Mexico) as well as English (US), for example), which gives you a second 100-character field for the same country. Apple does not document this. Treat it as a bonus you verify by tracking, not a rule you rely on.
The listing-level localization work (screenshots, video, the description itself) is covered in the ASO guide; this step is the keyword sets only.
Step 6: Track: read the ranks on a cadence
A keyword set is a hypothesis until you read the ranks. Read them on a cadence: weekly for the first month after any change, monthly after that, and always before you change anything again, so you can tell the change from the noise. Track your top twenty keepers, no more; a longer list stops getting read.
| What to read | Where | How often |
|---|---|---|
| Rank per keyword | Search the store yourself, or a tool; for iOS, Apple's public search API returns the ranked result list for a term and country (that is how the readings below were taken) | Weekly for a month, then monthly |
| Impressions and conversion | App Store Connect (App Analytics, search impressions and conversion rate); Play Console (store performance, search) | Monthly |
| Ratings count of the top three | The store search results for each keeper: your difficulty column, refreshed | Monthly |
| Which keyword to swap | Any keeper with no rank movement in eight weeks, or one that ranks but brings a conversion rate below your listing average | At each monthly read |
Swap one or two keywords per release, not the whole set; a full swap tells you nothing about what worked. Log every swap with its date in the sheet. If a phrase rises for four weeks and then falls, look at what changed on the top three competing listings before you touch yours. The app launch checklist has the first-week reads in order; the ranks here are the ones you keep reading after launch week.
First-hand: the workflow run on a published app
To make the six steps concrete we ran them on a customer's apps built and published with Appy Pie AI: CPA Reviewer on the App Store and CPALE Reviewer on Google Play, both by REO PH, a Philippine board-exam review provider. Everything below is public: the two listings, both stores' suggestions, and rank readings taken from Apple's search API and Google Play search on 24 September 2026. We did not change either listing; what the run produces is the recommendation the workflow would make, and we show it so you can judge the method.
Where the app ranks today
Apple's search API, US storefront, apps only, up to 200 results per term. The number after the rank is how many results the term returned; the app is either in that list at that position or it is not.
| Search term (iOS, US) | CPA Reviewer's rank | Results returned |
|---|---|---|
| cpa reviewer | 1 | 120 |
| cpa review | 5 | 126 |
| cpa study | 31 | 136 |
| far cpa | 38 | 122 |
| cpa practice questions | 48 | 142 |
| cpa exam | 56 | 152 |
| cpa exam prep | 110 | 153 |
| accounting exam | not in top 145 | 145 |
| Search term (Google Play, US) | CPALE Reviewer's rank |
|---|---|
| cpale reviewer | 1 |
| cpa reviewer | 10 |
| cpa exam | not in results |
What the run found
- The two names differ across stores. "CPA Reviewer" on iOS, "CPALE Reviewer" on Play. Play ranks it first for "cpale reviewer" because the word is in the name; iOS cannot, because on iOS the word appears nowhere in the indexed fields.
- The iOS subtitle is empty. That is 30 indexed characters, the second-strongest field Apple has, unused. Every rank in the iOS table is being earned by two words: "cpa" and "reviewer".
- The ranks follow the name exactly. First for "cpa reviewer", fifth for "cpa review" (the same stem, against Becker and UWorld with thousands of ratings), 56th for "cpa exam" and 110th for "cpa exam prep", where the phrase is not in the name and the competition is heavy.
- The best keyword on the sheet is the one the iOS listing lacks. "cpale reviewer" scored 11, the highest, because Apple suggests it and no app with real ratings owns it on the US storefront.
What the workflow recommends (not applied)
| iOS field | Today | Recommended | Why |
|---|---|---|---|
| Name (30) | CPA Reviewer | CPA Reviewer | Keep. It carries the brand and the two words that already rank. |
| Subtitle (30) | (empty) | CPALE board exam practice (25 chars) | Adds cpale, board, exam and practice, the top three keepers, with no word repeated from the name. |
| Keyword field (100) | Not visible to us | cpale,licensure,accountancy,accounting,auditing,taxation,far,afar,quiz,mock,questions,philippines (97 chars) | Single words, no plurals, nothing already in the name, subtitle or category. Apple combines them into the phrases on the sheet. |
On Google Play the recommendation is smaller: the name already carries the best keyword, and the short description uses 70 of its 80 characters well. The description is where the Maybe list ("cpa exam prep", "accounting exam") belongs, used in sentences, since Play has no keyword field to hold them. Nothing in either recommendation uses a competitor's name.
The apps in this run were published through Appy Pie AI's publishing service, which submits the listing fields for you; the keyword choices are still the developer's, which is the whole point of this guide.
Mobile app keywords beyond the stores
The scored list is worth more than the store fields it was built for. The same twenty phrases are what your app's landing page should rank for on Google, and the phrase people search in the store is very often the phrase they search on the web the day before. Put the top three keepers in the page title, the first heading and the first paragraph of the app's landing page; the landing page guide covers the rest of that page.
For paid acquisition the list splits. Apple Ads bids on App Store search terms directly, so your keepers are your bid list and your Maybe rows are your test budget. Google's App campaigns do not take keywords at all; they optimise on your listing text and creative, which is one more reason the Play description in step 4 has to be written well. The guide to getting users for your app puts both in the wider acquisition plan.
What are the most popular keywords on the App Store?
Nobody outside Apple knows, and any list that claims to is a model. Apple publishes no search volumes for the App Store. What it does publish, inside Apple Ads, is a search popularity score per keyword on a 5 to 100 scale, and every ASO tool builds its volume estimates from that score plus its own tracking. Those lists agree on the shape: the most searched terms are single generic words ("games", "music", "chat", "vpn") and big brand names, which is exactly why they are useless to a new app. A popular keyword you cannot rank for is worth less than an unpopular one you can own, which is the arithmetic the scoring template in step 3 is built around.
Use vendor popularity lists for one thing: finding the category phrases with the highest popularity that still have a beatable top three. Read them the way step 2 reads auto-suggest, as a signal of order, not as a count. The store comparison in our App Store versus Google Play statistics post shows how differently the two stores' users search, which is why this guide keeps the two lists separate.
Six keyword mistakes that cost rankings
Repeating words across iOS fields
A word in the name is already indexed. Repeating it in the subtitle or keyword field spends characters for nothing. Apple says so in its own guidance.
Chasing volume over relevance
A high-volume phrase you rank for that brings the wrong people lowers conversion, and both stores read conversion. Relevance is weighted double in the template for that reason.
Never revisiting
Most listings set keywords once. The competing top three change, the suggestions change, and a set that was right at launch is stale a year later. Step 6 exists to catch that.
Translating instead of localizing
A translated keyword field ranks for nothing because it is not what people type in that language. Localized sets are researched, not translated.
Ignoring the Play description
Play has no keyword field, so the description is doing the work the field does on iOS. A two-line description leaves most of the index empty.
Using competitor names
Grounds for rejection on iOS and a metadata-policy breach on Play. The names tell you who owns a phrase; they are not phrases you may use.
Frequently Asked Questions
How do I find keywords on Apple's App Store?
Type a two-word seed into the App Store search box on an iPhone and write down the suggestions in order; they are drawn from real searches and the order is a volume signal. Repeat for ten seeds, then mine the name, subtitle and description of the top ten listings for your best seed. Free tiers of AppTweak, AppFollow and Sensor Tower automate the mining and add a modelled volume. Step 2 above shows the suggestions we got for a real app.
What are good keywords for an app?
Phrases that describe what the app does for the person typing them, that the store suggests when you start typing, and where the top three results do not all have thousands of ratings. In the template above that is relevance 3, volume 2 or 3, difficulty 1 or 2, a score of 10 or more. A good keyword for a new app is usually two or three words long.
What are the most popular keywords on the App Store?
Apple does not publish search volumes, so nobody knows for certain. Apple Ads shows a search popularity score from 5 to 100, and ASO tools model volumes from it. The most popular terms are generic single words (games, music, chat) and brand names, which almost no app can rank for. Use popularity lists to find category phrases with a beatable top three, not to chase the head terms.
What are the four main types of keywords?
The classic web types are informational, navigational, commercial and transactional. In the stores they map to four intents: brand (your name and your competitors' names, which you may only use for your own), category ("budget app", "exam prep"), feature or problem ("receipt scanner", "practice questions") and modifier ("free", "offline", "for kids"). Category and feature phrases are where a new app should spend its characters.
How many keywords can I use on the App Store?
Apple gives you a 100-character keyword field, comma-separated with no spaces, plus a 30-character name and a 30-character subtitle, all indexed. That is roughly 12 to 15 single words in the field. Apple combines the words to match multi-word searches, so "board,exam" covers "board exam" without spending characters on the phrase.
Do keywords in the app description matter on iOS?
Apple lists the name, subtitle, keywords and category as what search matches against, and says promotional text does not affect ranking. The description is not on that list. Write the iOS description for people who have already found you; put the keywords in the three indexed fields. On Google Play it is the reverse: there is no keyword field and the description is indexed.
How long until a keyword change shows in the rankings?
On iOS the change ships with a new version, so it shows after review, usually within a day or two of approval, and settles over two to four weeks. On Google Play listing text can change without a release and re-indexes within days. Read ranks weekly for a month after any change before judging it; a single-day read is noise.
Are competitor names allowed as app store keywords?
No. Apple names "competing app names" and unauthorized trademarks as grounds for rejection, and Google Play's metadata policy covers unattributed use of other brands. Competitor names are useful in step 1 for finding the phrases they rank for; then they come off the sheet.
Chose Your Keywords? Publish the Listing Free.
Appy Pie AI's publishing service takes your name, subtitle, keywords and descriptions and submits the app to both stores for you. The keyword workflow above gives you the words; the publish page takes it from there.
Publish your app freeRead the full ASO guideKeywords Are the Cheapest Ranking Lever You Own. Use All the Characters.
The app in our run earns every one of its ranks from two words in its name, with thirty indexed characters empty beside them. Most listings look like that. Seed, expand, score, place, localize, track: six steps, one afternoon for the first run, twenty minutes a week after. If the app is not published yet, the ten-step guide to creating an app gets it there, and the listing fields are waiting at the publish page.
Publish your app free →
Aasif Khan
Head of SEO and Growth Marketing Lead at Appy Pie AI with 17+ years in digital marketing and scalable growth strategies.
Abhinav Girdhar
Founder and CEO of Appy Pie AI, one of the largest no-code and AI app platforms, with 10M+ apps and websites created.