App Store keyword research, step by step

You pick the words for your app name, subtitle and keyword field so that your app appears when people search for them. Apple says search results depend on text relevance (matches in your title, subtitle, keywords and primary category) and on user behavior such as downloads, ratings and reviews (Apple: App Store search).

You control the text. Good keyword research makes sure that text matches what people really type, in phrases where your app has a realistic chance to rank.

Start with a long, messy list. The best sources are the words other people use, not the words you use internally:

• Your App Store reviews and support emails: how do people describe the problem your app solves?
• Reviews of similar apps: which features do people ask for or miss?
• App Store search hints. Apple shows suggestions as soon as someone types in the search bar (Apple). Type your main word and write down what appears.
• The names and subtitles of apps that already rank for your main phrase.

Write phrases the way a person would type them, for example “habit tracker”, not “habit formation solution”.

Search every phrase and look at the top 10 apps. Two questions matter:

1. Are these apps like mine? If the top 10 is full of games and you make a budgeting app, the phrase means something else to Apple and to users. Drop it.
2. How strong are they? Rating counts are the clearest public signal. A top 10 where every app has hundreds of thousands of ratings is hard to break into. A top 10 with a few small apps is an opening.

Apple does not publish how often a phrase is searched, so every number you see in an ASO tool is an estimate. Use those numbers to compare phrases with each other, not as traffic forecasts. How LocalASO estimates volume and difficulty, with the exact formula, is in App Store keyword search volume and difficulty, explained.

Apple describes the trade-off itself: popular terms such as “jobs” or “social” can bring a lot of traffic but are highly competitive, while less common terms bring less traffic and less competition (Apple: product page).

A workable rule of thumb:

• Relevance first. A keyword that brings the wrong users hurts your conversion and does not stay.
• One or two broad phrases that describe the category, for the long run.
• Many specific phrases (two or three words) with medium demand and a top 10 you can beat. These bring the first downloads.
• Drop phrases nobody searches, even if you rank first for them.

You have three places for search text in each localization:

• App name (up to 30 characters) and subtitle (up to 30 characters), both visible on your product page (Apple: app information). See App Store title and subtitle: limits and examples.
• Keyword field (100, comma separated, invisible to users). See The App Store keyword field: limit, rules and examples.

Apple does not publish how much each field weighs. In practice most ASO people see the name count most, then the subtitle, then the keyword field, so put your most important phrase into the name or subtitle. Never repeat a word across the fields: Apple tells you not to repeat words from the name, subtitle or category in the keyword field (Apple). The free keyword field checker finds repeats for you.

Each storefront is its own search. People in Germany type “sprachen lernen”, people in Japan “語学学習” , and the apps that rank are different. A literal translation of your English keywords often misses the phrase locals actually use. Apple also recommends tailoring keywords to each market (Apple).

How to do this well, and which localizations each storefront offers, is in App Store localization for keywords. Country-specific notes are in App Store keywords by country.

The keyword field is part of your app version (Apple: platform version information), so changes go live with your next release. After a release, check your rank for each keyword every day for a couple of weeks and judge the trend, not a single day. Replace keywords that bring nothing with the next candidates from your list.

How to check your App Store ranking by country explains how ranks are measured.

LocalASO does steps 2 to 7 on your Mac: it shows your keywords as rows and countries as columns, with your rank, estimated volume and difficulty in each cell, adapts each keyword to the local phrase for every country with the AI you already use, and suggests a name, subtitle and keyword field per language. See the keyword research docs for the details.