App Store Keyword Research: Finding Terms You Can Actually Rank For
How to do App Store and Google Play keyword research properly: where to find candidates, how to read volume and difficulty, and what to target.
· · by the ASOnest team
Good app keyword research is a filtering problem, not a discovery problem. Generate candidates from four sources — autocomplete, competitors, reviews and semantic expansion — then filter hard on difficulty relative to your own app's authority. A term you can rank #3 for beats a term with ten times the volume where you rank #47.
The hard part of app keyword research is not finding keywords. Any tool will hand you four thousand of them in ten seconds. The hard part is throwing away three thousand nine hundred and sixty of them correctly.
Why volume-first research fails
Open any keyword tool, sort by search volume descending, and the top of that list is a graveyard. Those terms are contested by the largest apps in the category — the ones with hundreds of thousands of ratings, years of download history and marketing budgets.
Ranking is not a participation trophy. Position 47 for a high-volume term produces approximately nothing. App store search traffic collapses steeply with position: the top three results take the overwhelming majority of taps, and by the second screen you are functionally invisible.
So the question is never “which keywords have the most searches?” It is “which keywords have enough searches, and a top ten I can plausibly join?”
The four sources of good candidates
1. Store autocomplete
Autocomplete reflects what people actually type, including the misspellings and the oddly specific phrasings. Type your core noun and record every suggestion, then repeat with each letter of the alphabet appended.
Do it per country. torch in the UK and flashlight in the US are the same app and different keyword universes.
2. Competitor keyword sets
This is the highest-yield source and the most underused.
Find the five apps that keep appearing next to yours in search results. Not the apps you consider competitors commercially — the apps that share your search results, which is often a different list. Then take every keyword they rank in the top twenty for.
A competitor at position #4 for a term has proved two useful things at once: the term produces installs, and a top-ten slot is achievable for an app of roughly their size. If their rating count is within an order of magnitude of yours, that keyword is a candidate.
3. Your own reviews
Your users describe your app in language you would never choose. You wrote “meal planning app”. Two hundred reviewers wrote “weekly dinner organiser”, “what to cook this week”, “family menu app”.
Those phrases carry high purchase intent because they come from people who already found the category valuable. Pull recurring phrases out of your reviews, check each for real search volume, and keep the ones with demand.
4. Semantic expansion
The terms adjacent to your seed list that you did not think of. A meal planner sits near grocery lists, pantry trackers, batch cooking and macro tracking — different intents that partially overlap with your app.
Expansion is worth doing, but it produces the most noise of the four sources. Filter it hardest.
Reading volume properly
Apple exposes a search popularity score from 0 to 100 through Apple Ads. It is logarithmic, not linear — a score of 60 is not twice the traffic of 30, it is many times more. Most useful long-tail terms for a small app sit between 5 and 35.
Google Play does not publish an equivalent — its search guidance is qualitative — so Play volume is inferred from demand signals and autocomplete frequency. Treat Play figures as a reliable ordering and a rough magnitude rather than a precise count.
Practical thresholds:
| Popularity | What it means for an indie app |
|---|---|
| Below 5 | Effectively no traffic. Skip unless it is exactly your niche. |
| 5–25 | The sweet spot. Real traffic, usually reachable. |
| 25–50 | Worth targeting once you have some rating volume. |
| Above 50 | Almost always contested by very large apps. |
Reading difficulty properly
Difficulty answers: how hard is it to break into the top ten for this term?
The inputs that matter:
- Rating counts of the current top ten. The dominant factor. Ten apps averaging 200,000 ratings is a wall.
- Rating velocity. An app gaining 5,000 ratings a month is harder to displace than one at the same total that has stalled.
- Keyword prominence in their metadata. An exact-match title is the strongest possible signal and the hardest to beat.
- Top-ten stability. If the same ten apps have held those positions for ninety days, the term is frozen.
Difficulty is not the same as winnability
This distinction is where most tools stop and where most decisions actually get made.
Difficulty describes the keyword in isolation. Winnability describes the keyword relative to your app — your rating count, your category position, your current metadata strength, measured against the apps in that specific top ten.
Two apps looking at the same difficulty-45 keyword get very different answers. An app with 40,000 ratings should target it. An app with 300 should not, and no amount of clever copywriting changes that.
A worked example
A meal planning app with about 800 ratings, six months old. Candidate list after filtering:
| Keyword | Popularity | Difficulty | Verdict |
|---|---|---|---|
| recipe app | 72 | 88 | Skip — owned by apps 100× larger |
| meal planner | 61 | 34 | Target — surprisingly soft top ten |
| recipe organizer | 38 | 21 | Target — clear opportunity |
| grocery list maker | 29 | 18 | Target — easy win |
| weekly menu planner | 24 | 16 | Target — likely top three |
| cookbook app | 31 | 55 | Hold — revisit at 3,000 ratings |
Four targets, one to revisit, one to abandon. That is the correct output of a keyword research session: a short list you can act on within your character budget.
Note that recipe app — the term the founder would have picked by instinct — is the one to drop.
Placing the winners
Once you have your list, ordering matters:
- Highest-value phrase → app name. The name carries the most ranking weight, and it is where you put the one term you most want.
- Second cluster → subtitle. A different theme, never a restatement. Repeating a word wastes the field.
- Everything else → keyword field. Comma-separated, no spaces, singular forms, no word that already appears in the name or subtitle.
Remember that Apple constructs combinations across fields. meal in the keyword field and planner in the subtitle makes you eligible for meal planner without spending characters on the phrase. Planning around combinations is where the last twenty per cent of coverage comes from.
Google Play works differently — no keyword field, and the long description is indexed — so the same research feeds a different placement strategy. That difference is covered in App Store vs Google Play ASO.
How often to redo this
Quarterly is about right, plus:
- After a significant feature launch that changes what your app is
- When you enter a new country
- When a tracked competitor makes a large metadata change
- When your rating count grows by roughly an order of magnitude — this genuinely changes which keywords are winnable
Between those moments, hold. Every rewrite resets your measurement baseline, and a listing that changes monthly never accumulates enough data to tell you anything.
Doing this by hand is possible but slow — the volume and difficulty numbers are the parts you cannot eyeball. ASOnest scores volume, difficulty and winnability on both stores, across all countries. Every plan starts with a 7-day trial and no card.
Frequently asked questions
There is no universal threshold, because volume only means something next to difficulty. A popularity score of 20 with difficulty 15 is usually a better target for a small app than a score of 70 with difficulty 85. As a rough guide, terms below about 5 popularity on Apple's scale attract too little traffic to matter, and terms above 60 are almost always contested by apps with very large rating counts.
Between fifteen and forty for a single app. You have roughly 160 usable characters on the App Store, and Apple builds combinations across your fields, so a well-constructed set of fields makes you eligible for far more terms than you literally wrote. Tracking more than about forty means most of them get no attention.
Yes, in two ways. Apple constructs combinations across your name, subtitle and keyword field, so 'meal' and 'planner' in separate fields make you eligible for 'meal planner'. Both stores also associate apps with terms based on user behaviour and category context. But deliberate placement beats accident every time.