How to Do ASO for an App: A Step-by-Step Workflow
A repeatable ASO workflow for one app and one market: gather competitor phrases, expand with store autocomplete, score, write the listing, ship, and check rank.
**How to do ASO** is mostly a sequencing problem. Most teams start with a title rewrite, skip the research, and then have no idea whether the new title helped. A better workflow starts with evidence, narrows to a small set of phrases, and changes the listing in an order you can measure.
This is the process NewASO is built around, written so you can run it with or without the tool.
Key takeaways
- Start with one app, one store, and one country. Broad research produces broad, unusable lists.
- Mine competitors first; they have already done the market's keyword discovery.
- Use store autocomplete to confirm real phrasing, not a generic keyword database.
- Score before writing. A long list becomes useful only when it is ranked.
- Change one listing layer at a time and check rank on demand.
Step 1: Define the unit of work
ASO gets vague when the scope is vague. Choose one app, one storefront, and one store. If the app is live in twelve countries, start with the country that produces the most revenue or the one you understand best. Research one market at a time so the scores stay comparable.
Write down what the app actually does in one sentence, without adjectives. That sentence is your relevance filter for every phrase that follows. If a keyword requires a claim the product cannot support, it is not a candidate, no matter how attractive the number looks.
Step 2: Mine the competitors
Find three to five apps that rank for the obvious phrases in your niche. Read their titles, subtitles, and descriptions. Pull out the two-to-four-word phrases that appear across more than one competitor. Repetition is a cheap signal that the phrase matters in the category, and it gives you roots rather than random guesses.
Keep a note of the phrases each competitor repeats. A phrase owned by one giant may be a dead end for a new app; a phrase repeated by several mid-sized apps is often a more realistic target. NewASO separates these into newcomer keywords and phrases giants own, which is a useful distinction even if you do the work by hand.
Step 3: Expand with real store suggestions
Feed each root into Google Play and App Store search. Note the autocomplete suggestions and the apps that appear. Autocomplete is not a volume oracle, but it is direct evidence that people type the phrase into that store. Expand with a few synonyms and adjacent intents, then deduplicate.
- Keep phrases of two or three words; single words are usually too broad to win.
- Reject phrases that describe a different job to be done, even if they share a word.
- Save the source of each phrase so you can tell a competitor root from a synonym later.
- Do not translate a list from another language and call it localisation; phrasing changes.
Step 4: Score, then choose a small set
Score each phrase on demand, competition, difficulty, relevance, and opportunity. You do not need a perfect formula. You need a consistent one that forces you to compare phrases instead of falling for the biggest number. NewASO uses a 0–100 scale and a confidence label; the opportunity score is relevance multiplied by volume and reduced by difficulty.
| Band | What to do with it | Where it belongs |
|---|---|---|
| Focus | Best combination of fit and winnability | Title, subtitle, or first description line |
| Primary | Strong supporting demand | Subtitle, keyword field, short description |
| Secondary | Useful but not differentiating | Full description, later screenshots |
| Maybe | Needs more evidence or a different market | Watchlist |
| Skip | Wrong intent, too competitive, or misleading | Nowhere |
Step 5: Write the listing in store order
Start with the title and subtitle or short description. These are the fields a user sees in search and the fields that carry the most weight. Then write the first three screenshots as a single argument. Only then write the long description for the people who are already interested. Check every field against the store's character limit and banned-word rules before you paste it.
Avoid stuffing. A title with five keywords separated by dashes reads like spam and converts worse. The phrase should sound like the name of a product a person would install.
Step 6: Ship, then check rank on demand
After the update is live, run a rank check for the phrases you targeted. Do not assume the store applied your changes immediately, and do not trust a dashboard that claims daily precision without saying where the number came from. Check again after a reasonable interval, then compare.
If a phrase did not move, the cause is usually one of four things: the field did not index it, the competition is stronger than the score suggested, the listing does not convert well enough for that query, or the app is too new to compete. Diagnose before rewriting everything.
Where teams get stuck
The common failure is not a lack of keywords. It is a lack of decisions. A list of 1,200 phrases feels productive, but the work only starts when you choose twenty, write for five, and measure three. The other common failure is changing five things at once, which makes the result unreadable.
NewASO handles the mechanical parts: competitor mining, autocomplete expansion, scoring, field validation, localisation, and rank checks. The judgement is still yours. If you want to see how those pieces fit, browse the feature list or read the ASO checklist before you start.
Frequently asked questions
What should I do first for ASO?
Pick one app, one store, and one country. Mine three competitors for repeated phrases, expand those roots with store autocomplete, and score the result. Do that before changing any copy.
How many keywords should I target?
Aim for a shortlist of twenty to forty scored phrases, with five to ten that clearly belong in high-weight fields. A longer list is fine for research, but it is not a plan until it is ranked and pruned.
Should I update the listing every week?
No. Ship meaningful changes and give them time to collect impressions. Constant churn makes it impossible to attribute a rank or conversion change. A monthly research and update cadence is usually enough for a small app.