App Store Keyword Research: A Playbook for Indie Developers (2026)

ASO keyword research is the one ASO task that can consume unlimited time. Paid tools charge $99 to $400 per month to automate the work. This post covers the zero-budget version — what an indie developer actually does, in order, with App Store autocomplete, Top Charts, and a character counter as the entire research stack.

The short version: the tools are selling automation, not unique data. Every paid tracker is querying the same App Store autocomplete endpoint and the same public Top Charts. At indie scale you can do the same research by hand in an afternoon, and the signal quality is indistinguishable until you’re past roughly 10,000 installs.

The workflow below is five steps executed in strict order. Pair it with the ASO playbook for indie developers for the broader context, and with the Apple App Store keywords field deep-dive for the mechanics of the 100-character field.

What you are actually researching

Before you open the App Store app, know the shape of the output. A complete keyword research pass produces three tiers.

• Primary keywords (1–3): the terms that define your app. These go in the title, subtitle, and first slot of the keywords field.
• Secondary keywords (5–10):terms that add search surface but aren’t the core pitch. These fill the remaining keywords field characters.
• Long-tail queries (15+):multi-word combinations Apple constructs automatically from your primary and secondary terms. You don’t type these into the keywords field — Apple builds them for you by combining words across title, subtitle, and keywords field.

What Apple indexes for ranking: title + subtitle + keywords field + in-app purchase names + developer name. The description is not indexed for iOS ranking.

What Google indexes for ranking: title + short description + full description. No dedicated keywords field. Covered in the App Store vs Google Play listing requirements post.

The 5-step indie keyword research workflow

Five steps, executed in order. Each feeds the next.

Step 1: Define your app in one sentence

Write one sentence that describes what your app does and who it’s for. No brand name, no marketing. “An offline-first markdown notes app for writers.” “A budget tracker for couples sharing household expenses.” “A sleep timer for parents of infants.” If you can’t define it in a sentence, keyword research is premature — come back when the product is sharper.

Step 2: List 20 candidate keywords from that sentence

Brainstorm mode, no filtering. Pull out category-level terms, benefit-level terms, audience-level terms, verb-level terms. For a markdown notes app: markdown, notes, journal, outline, bullet, write, writing, writer, draft, editor, offline, sync, minimalist, focus, distraction-free, essay, plain text, research, daily, diary.

Step 3: Verify with App Store autocomplete

Open the App Store app on your phone, tap Search, and type each seed keyword one letter at a time. Record the suggestions Apple returns. Apple ranks autocomplete suggestions by real search volume, so the first 3–5 suggestions per seed are the terms with meaningful traffic. If your candidate keyword doesn’t appear in any autocomplete, nobody is searching for it — drop it.

Step 4: Filter with competitor check

Open the top 30 apps in your App Store category. Read each app’s title, subtitle, and first description sentence. Build a frequency map: how many times does each of your candidate keywords appear? If five or more competitors already use a keyword, it’s likely dominated — you probably can’t rank for it in 90 days. Strike those. Keep the keywords that appear in 1–3 competitors (unowned or barely owned) or in zero (unoccupied territory).

Step 5: Prioritise to three keywords you can own

From the filtered list, pick the three highest-intent, lowest-competition keywords that genuinely describe your app. Apply the three-keyword rule (below). Drop the other 17. Those three become the centre of your metadata; everything else is secondary.

The three-keyword rule

The principle: Apple’s keywords field is 100 characters. Your three highest-priority terms should occupy the title, the subtitle, and the first slot of the keywords field — the three highest-weight ranking positions. Everything else is supporting cast.

Why three and not thirty?Two reasons. First, attention. You cannot iterate meaningfully on 30 keywords across 30-day release cycles; you can iterate on three. Second, evidence. If you cannot rank in the top 10 for at least one of your three keywords within 90 days, the research failed — you picked terms you can’t own. Re-run the workflow with easier candidates.

Picking criteria. Each of your three keywords must pass three tests:

1. Intent match. Does this keyword describe something a user who wants your app would search?
2. Realistic difficulty. Can you beat the 20th- to 30th-ranked app in the category for this term?
3. Product fit.Can your first screenshot and subtitle honestly sell this keyword’s promise in 2.5 seconds?

Use the free keyword character counter to validate the 100-char limit and the subtitle helper to craft a subtitle that carries one of your three keywords cleanly.

Free research sources that actually work

Six sources, ranked by signal-to-effort ratio. The first three cover most indie apps; the next three fill gaps.

1. App Store search autocomplete. The highest-signal free source. Apple ranks suggestions by real search volume, and indie-scale apps get the same access to this data that paid tools scrape.
2. App Store Top Charts.Open your app’s category Top Charts on the App Store app. Read the top 30 apps’ metadata. This is direct competitor signal and free.
3. Google Play autocomplete.Same idea as Apple’s but on the Play Store app. Critical for Android-first or cross-platform apps, since Google indexes keywords very differently from Apple.
4. Apple Search Ads Basic.The free tier of Apple Search Ads surfaces Apple’s own suggested keyword list for your app. You don’t have to run any ad campaign — just create the account, add your app, and read the suggested-keyword panel.
5. Product Hunt + Indie Hackers topic pages. What users describe their needs with, in their own words. Surfaces keyword variants App Store search won’t show you.
6. Reddit subreddit searches. Use site:reddit.com in Google search plus your seed keyword. Reddit users describe problems in natural language that often maps to underexplored long-tail keywords.

Paid alternatives— ASO.dev, App Radar, AppTweak, Sensor Tower — are worth considering once you’re past roughly 10,000 installs and need cross-market ranking tracking and daily keyword-position history. Below that scale the automation overhead exceeds the signal gain.

Using App Store autocomplete as your data source

Autocomplete deserves its own walkthrough because it is the single most valuable free keyword research tool that exists, and most indies never use it systematically.

The procedure:

1. Open the App Store app on your phone, not the desktop Mac App Store. Autocomplete data differs between the two surfaces; mobile is the signal you want.
2. Tap the Search tab.
3. Type each seed keyword one letter at a time, and record the suggestions Apple returns at each prefix. The full prefix (e.g. typing “markdow” through “markdown”) reveals autocomplete variants that only show at certain character counts.
4. Record the first 3–5 suggestions per seed in a spreadsheet, with the rank position as a volume proxy. Higher position = more search volume.
5. Repeat for each locale you plan to ship. Change the App Store country in your Apple ID settings — each region returns different autocomplete data. Japanese autocomplete has nothing to do with English autocomplete; research each locale independently (see localization post).

Accelerating the workflow with AI

AI is excellent at step 2 (candidate generation) and step 5 (filtering against the 100-character budget). It is not a replacement for the autocomplete verification in step 3.

Push My App’s AI metadata generator takes a plain-English app description and outputs:

• 10+ keyword candidates, each with rationale.
• An estimated difficulty score per keyword.
• A 100-character keywords field string that fits the candidates respecting Apple’s field limit and avoiding duplicates with the title and subtitle.
• Per-locale keyword regeneration for each of 14 locales — researched per locale, not translated.

Use AI as step 2b: expand your 20-candidate brainstorm list to 30–40 candidates. Then filter with the autocomplete + competitor pass as before. AI generation + manual verification beats either alone — AI without verification produces plausible-looking keywords that nobody searches for; manual without AI misses synonyms and phrasings you would not have thought of.

Pitfalls that waste research time

Five mistakes that turn a research pass into a busywork pass.

Chasing high-volume category terms you cannot rank for. Photo editor, todo app, budget tracker — dominated by incumbents with years of install history. Pick niche terms you can actually own.

Ignoring intent. Productivity has volume; it has weak install intent for a niche app. A query that leads to a more specific follow-up query (e.g. productivity → markdown editor) is a browse query, not an install query.

Not testing against App Store autocomplete. The most common time-waster. A keyword that seems obviously relevant but doesn’t appear in autocomplete has negligible search volume — the App Store tells you this for free.

Assuming paid tools know better. For niche apps, paid tools often track generic categories and miss your specific intersection. At indie scale, your manual research is at least as good as any paid tracker.

Translating keywords for localization instead of re-researching.Japanese users don’t search the Japanese translation of task; they search transliterated todo. Every locale needs its own research pass. Covered in the How to Localize an iOS App in 14 Languages post. And before submitting, run the full ASO pre-submission checklist.

Frequently asked questions