Guide
App Store Keyword Research: A Step-by-Step Guide for 2026
App Store keyword research means listing the words people type, scoring each for demand and difficulty per country, and placing the winners in your metadata.
App Store keyword research is the process of listing the words people type into App Store search, scoring each one for demand and difficulty in each country, and putting the best ones into your app name, subtitle and keyword field. Done well, it gives you a short list of keywords you can realistically rank in the top 10 for, instead of a long list of popular words you’ll never be seen for.
This guide walks through the process step by step, with a worked example for a budgeting app, a decision rule you can reuse, and the limits of every data source.
What Apple Actually Uses
Before picking words, know where they go. Apple’s App Store search page says results depend on “text relevance (matches for your app’s title, subtitle, keywords, and primary category), as well as user behavior (downloads, ratings and reviews, and more).”
So keyword research feeds four places:
| Field | Limit | Source |
|---|---|---|
| App name | 30 characters | App Store Connect: App information |
| Subtitle | 30 characters | Same page |
| Keyword field | 100 characters, commas, no spaces | Apple: Product page |
| Primary and secondary category | Chosen from Apple’s list | Apple: App Store search |
The description and promotional text don’t count. Apple says plainly that promotional text “doesn’t affect your app’s search ranking.”
Step 1: Write the Jobs Your App Does
Start from the user’s words, not your feature list. Write three to five jobs your app does, phrased the way someone would search:
- “budget planner”
- “track spending”
- “bill reminder”
- “savings goal”
If you can’t phrase a job in two or three plain words, people probably don’t search for it.
Step 2: Expand Into Candidates
Turn each job into 10 to 30 candidates. Good sources, in order of usefulness:
- App Store search suggestions. Type the start of a keyword on an iPhone set to the country you’re researching and note the suggestions. They reflect what people type there.
- Competitors’ names and subtitles. The apps already in the top 10 tell you which words Apple associates with the job. Note their words, never their brand names: Apple’s App Review Guideline 2.3.7 forbids packing metadata with “popular app names.”
- Your reviews. Users describe your app in their own words. “Finally a simple expense log” is a keyword idea.
- Synonyms and regional words. “Paycheck” in the US, “payday” in the UK.
Aim for 50 to 150 candidates per country. Our App Store keyword research tool page shows how AppMeerkat suggests candidates from the store itself.
Step 3: Score Demand and Difficulty, per Country
Every candidate needs two numbers and one look.
Demand. Apple doesn’t publish search counts. The closest official signal is the Popularity Apple Ads shows when you research keywords for a campaign. It’s relative, coarse at the low end, and only available inside Apple Ads. Our guide to Apple’s search popularity score explains how to read it. Third-party tools, including AppMeerkat, add estimates; AppMeerkat labels its own as Demand (Est.).
Difficulty. No tool knows Apple’s ranking formula, so difficulty is always an estimate. The practical proxy is the live top 10:
- How many of the top 10 have the keyword in their name or subtitle?
- How many ratings does the weakest app in the top 10 have, in this country?
- Are the top results big brands, or apps like yours?
The look. Search the keyword yourself in that country. If the results are about something else (for example “goal” returns football apps), the keyword isn’t relevant, whatever its demand.
Step 4: Apply a Decision Rule
Here is a rule you can apply to every candidate. It’s deliberately simple, so you apply it the same way each time.
| Relevance | Demand | Top 10 you can join? | Decision |
|---|---|---|---|
| High | Any above the floor | Yes | Target now (name, subtitle or keyword field) |
| High | At the floor | Yes | Keyword field filler if space remains |
| High | High | No | Long-term; target a longer phrase containing it |
| Low | Any | Any | Drop |
“Top 10 you can join” means: the weakest app in the top 10 has a rating count within reach of yours in that country, and at least a few results don’t have the exact keyword in their name.
Worked Example: a Budgeting App in the United States
Our example app is “Penny”, a hypothetical budgeting app with about 300 US ratings. These are the notes for six candidates. The demand and difficulty labels are illustrative, to show how the rule works; your own numbers will differ.
| Keyword | Relevance | Demand | Weakest top-10 app | Decision |
|---|---|---|---|---|
| budget | High | High | Tens of thousands of ratings | Long-term |
| budget planner | High | Medium | A few thousand ratings | Target now |
| bill reminder | High | Medium | Several hundred ratings | Target now |
| paycheck budget | High | Low | Under 100 ratings | Keyword field |
| goal | Low | High | Sports apps | Drop |
| couple budget | High | At the floor | Under 50 ratings | Keyword field |
“Budget” alone is out of reach for a 300-rating app, but “budget planner” contains it. Putting “Budget Planner” in the name targets both the phrase and, over time, the single word.
Step 5: Place Keywords in Your Metadata
Placement order:
- Name (30 characters): your brand plus the strongest relevant keyword. Penny’s becomes “Penny: Budget Planner” (21 characters).
- Subtitle (30 characters): the next most valuable phrase that reads naturally. “Expense Tracker & Bill Alerts” (29 characters).
- Keyword field (100 characters): single words that combine with the name and subtitle. Don’t repeat any word already used; Apple says not to repeat “words included in your app name, subtitle, or category.”
Penny’s keyword field, at 97 characters:
money,spending,saving,debt,paycheck,household,monthly,allowance,receipt,goal,wallet,salary,family
You may notice “goal” is here even though the decision table dropped it. As a keyword on its own, its results are about sports. As a word, it lets “saving” and “goal” combine into “savings goal”, a phrase Penny does target. That’s the difference between a keyword you target alone and a word you include so it combines. Our keyword field guide covers the rules in detail, and the free keyword field optimizer checks length, duplicates and plurals for you.
Step 6: Track Ranks and Iterate
Keyword changes go live only with a new app version, so each round is slow. Make each round count:
- Record ranks for every targeted keyword daily, starting a few days before the update, so you have a baseline.
- Mark the release date on your history.
- Wait two to four weeks before judging. Ranks often jump around for days after an update.
- Swap out keywords that didn’t enter the top 50 for ones from your “keyword field filler” list.
Track every country you care about separately; ranks for the same English keyword differ between the US, the UK and Australia. Our guide on tracking keyword rankings in every country covers the cadence. AppMeerkat checks ranks daily in every country the store serves and keeps the full history, so you can see what each change did.
Limits and When Not to Bother
- No one has Apple’s real search counts. Popularity is relative and coarse; every other number is an estimate. Treat scores as a way to sort candidates, not as forecasts.
- Ranking well doesn’t guarantee downloads. A rank 3 for an irrelevant keyword brings views and no installs, and that can hurt you.
- Brand-new apps with zero ratings should target long, specific phrases first. Broad words are a later step.
- If your app is found mainly through word of mouth or ads, keyword research still helps, but the gain is smaller. Spend the time on screenshots and ratings instead.
Checklist
- Three to five jobs written in user words
- 50 to 150 candidates per main country
- Demand noted, with estimates labelled as estimates
- Live top 10 checked for each shortlisted keyword
- Decision rule applied the same way to every keyword
- Name and subtitle under 30 characters each
- Keyword field under 100 characters, no repeated words, no competitor names
- Baseline ranks recorded before release; release date marked
AppMeerkat is launching soon; if you want daily ranks and scores per country on your own computer, get notified at launch.
Questions
How many keywords should I research for one app?
Start with 50 to 150 candidates per country. You will only fit 10 to 20 of them in your name, subtitle and 100-character keyword field, but you need the wider list to find the ones you can actually rank for.
Where can I see real App Store search volume?
Apple doesn't publish search counts. The closest official signal is the search popularity Apple Ads shows for keywords. Every third-party number, including AppMeerkat's Demand (Est.), is an estimate.
Should I research keywords separately for each country?
Yes. The same English word can be popular in the United States and rare in the United Kingdom, and the top 10 apps differ by country. Research your main countries one at a time.
How often should I redo keyword research?
Review it with every app update that changes metadata, and do a full pass every two to three months. Keyword changes on the App Store need a new app version, so batch them.
Read next
Launching soon
Let a meerkat watch your ranks.
AppMeerkat tracks every keyword, in every country, on both stores. One email at launch, with the early-bird price.