How to choose the keywords to target for a new app

This guide takes you from a blank keyword list to a shortlist of terms your new app can realistically compete for, tracked and ready to go into your listing. It needs a tracked app, and nothing here requires you to have ranked for anything yet. Everything runs on the free plan, whose monthly inspections and autocomplete queries cover a first shortlist and whose five keyword slots hold the terms you keep (see pricing).

Choosing the keywords for a new app, the short version:

1. Collect seeds, starting from what Applyra already found for your app.
2. Turn them into real queries with the Autocomplete Simulator.
3. Inspect each candidate: Easy or soft Medium, with real demand.
4. Rank the survivors by KEI.
5. Check them against your weight class in the Metadata Simulator.
6. Track the keepers: one brand term, the rest winnable.
7. Put them in your listing, then read two weeks of positions.

1. Collect the seeds

Start from the words someone would type if they did not know your app existed. Not your category, the job. A habit app is not "productivity", it is "habit tracker", "daily routine", "streak". A game is a compound of mechanic, theme and audience, "word puzzle for adults", not "puzzle": in the mobile game ASO data, single genre nouns are the most contested terms and small games reach the top 10 on them far less often.

Before you invent anything, three places in Applyra already hand you candidates with their scores attached:

• Keywords to track, on your app page: search terms taken from your own listing that you do not follow yet, each with Traffic, Difficulty and KEI.
• Niche Analysis, in the Tools group: type the topic of your app and it returns the sub-niches inside it, each with its keywords, their traffic and difficulty, and the apps already there.
• The Head-to-head tab of a competitor: the keywords they rank on and you do not, under They lead, and the ones neither of you holds yet, under Neither. See head to head.

What good looks like: 10 to 20 seeds, most of them two words, none of them your brand name.

2. Turn seeds into real queries

The Autocomplete Simulator, in the Tools group, asks the store itself what it proposes as you type. Pick the store, country and language, type a prefix a letter at a time, and you get the store's own suggestion list.

A seed you invented and a query people type are different things. If the store completes "habit" into "habit tracker" and "habit tracker free" but never into your phrasing, your phrasing has no demand behind it. Everything the store offers back is something someone searched. Each suggestion carries an Inspect button that opens it in the Keywords Inspector on the same market, so the next step is one click away.

What good looks like: 20 to 25 candidate phrases, all of them returned by the store rather than invented by you.

3. Inspect each candidate

The Keywords Inspector, in the Tools group, takes one keyword in one market and returns its two scores, its KEI verdict, a breakdown of the competition and today's top 100.

Under Competition Analysis, Competitor Strength and Competitor Breadth say how hard the apps holding the term are to dislodge, Autocomplete Demand and Market Size how much demand sits behind it, and Key Observations puts it in plain words, including whether the competition thins out past the leaders or stays strong all the way down.

Two more sources of candidates sit on the same page. The Related Keywords card, next to Tracking Status, lists what the store suggests around the term: a head term you reject usually has a two word cousin that is easier, as long as it still carries traffic. And the Most Repeating Words strip above the Top Apps table shows the words the ranking apps put in their titles.

Everything you inspect stays in your Inspection history, under the search form, with its current scores. That table is your shortlist; you do not need a text file.

What good looks like: Easy or the soft end of Medium on difficulty, and a traffic score of 25 or more. Terms that clear both are rarer than they sound, as the State of ASO 2026 measured across the keywords indie developers track.

4. Rank the survivors by KEI

KEI weighs demand against competition in a single number, so it settles the argument between a quiet easy term and a busier medium one without you having to have an opinion. The inspector shows it at the top of Competition Analysis, as a verdict badge with the number beside it.

Keep what Applyra rates Good or Excellent, 0.6 and above, and order it from the highest KEI down: that is the order the terms go into your listing. See KEI.

What good looks like: a ranked shortlist of terms rated Good or Excellent. On the free plan, the top four plus your brand are the ones you will track.

5. Check them against your weight class

A new app is not in the same race as a five year old one. The Targeting fit axis of ASO Health grades the terms in a listing for an app at your stage, and you can run it on your shortlist before any of it is live.

On your app page, open the ASO Health tab and press Simulate changes. In the Metadata Simulator, write your top terms into a draft title, subtitle or short description and keyword field, and press Score this listing. The Targeting fit detail marks each term Within reach, Ambitious or Out of reach, and raises Your title aims above your weight class when the title carries a term the category giants hold.

Take an Out of reach verdict as arithmetic, not criticism: those characters would rank nowhere. Swap the term for a more specific variation, and revisit it later, because Targeting fit grows more forgiving as your app matures.

What good looks like: at most one or two Ambitious terms, and nothing Out of reach in the title.

6. Track the keepers: one brand term, the rest winnable

Track each keeper from the Keywords Inspector with Track keyword, or all at once from the Keywords Tracker, in the App Store Optimization group, with Add keywords. On the free plan that is five: one brand term, a control you expect to hold near the top, and four winnable terms. See plans and limits.

Applyra measures each keyword's first position the moment you add it. A dash in the Position column means your app is outside the top 100 today, which is a measurement, not a gap. From then on each row carries Position, Traffic, Difficulty, KEI, Trend (7d), and the apps ranked directly Ahead and Behind you.

7. Put them in your listing, then read two weeks of positions

Tracking a term records where you stand; it does not move anything. The positions move once the terms are in your listing, which is the next guide: optimizing an App Store listing or optimizing a Google Play listing.

The store is fast. In our study of real metadata changes, a keyword entering the title made most of its gain by the first daily reading. What takes weeks is reading the result: track twenty keywords and, on a completely uneventful day, one or two move five places or more on their own. So give it two weeks of daily readings, judge the whole set rather than one term, then decide. The free plan keeps seven days of history, enough to see the first move; reading the two weeks on one chart takes Unlimited.