THE ASO PROFESSIONAL FIELD GUIDE / App store keyword search intent
READ THE NEED. NOT JUST THE QUERY.
A query is a small piece of evidence about a person’s need. Your job is to interpret that evidence carefully, then decide whether your app can satisfy the expectation.
Search intent is the task or expectation behind a query. For ASO, interpret that expectation, check whether the product fulfils it and make the listing demonstrate the fit. Shared words do not necessarily mean shared intent.
Your deliverable: Separate similar-looking searches by the actual job the user expects the app to perform.
Fictional examples and editorial frameworks. No live keyword data, client result or ranking forecast is shown.
01
Translate the phrase into a user task.
Write the expected action after installation in a plain sentence. Then ask whether a different reasonable interpretation would require a different product or message. Preserve those alternatives in the record until evidence helps distinguish them.
The PaceStart assessment below uses running queries to show how audience cues, supported workflows and ambiguous intent should be recorded together.
02
Add the context you actually know.
Specify the market, language and store. Review relevant results and product expectations to understand how the phrase is used. Do not assume an English phrase means the same thing in every market or that a translation preserves the original intent.
Use seed research to collect candidate language and sources. Treat a single observed result page as a snapshot, not a complete picture of every user’s experience.
03
Separate audience fit from linguistic similarity.
Two phrases can share a word and still imply different products. A “running tracker” and a “running training plan” may overlap, but one emphasises recording activity while the other promises guidance. Your listing must support the promise it attracts.
The relevance guide provides a product-based check. Similar words are not enough to justify putting unrelated needs in the same priority group.
04
Connect the interpretation to the listing.
If the chosen audience wants a weekly meal plan, the opening copy and screenshots should demonstrate that experience. An advanced nutrition dashboard may be valuable later, but it does not necessarily answer the first question that brought the visitor there.
Use screenshot strategy to select the evidence. The illustrative map shows this connection; it is not a claim that a particular message produces a measured uplift.
05
Test your interpretation against new evidence.
Review how actual users describe the product and where the listing still creates confusion. Search intent is an interpretation that can improve as evidence accumulates. Keep a note of what would change your conclusion.
Once the groups are clear, use prioritisation to decide which deserves attention. Do not let a demand estimate override a product mismatch simply because the number looks attractive.
WORKED EXAMPLE / FICTIONAL APP
Three meanings hiding behind a running keyword
PaceStart is a fictional beginner running app. The team groups “running routes,” “running tracker” and “running plan” together because each contains running. Yet one person wants somewhere to run, another wants a record of a run, and another wants help deciding what to do tomorrow. PaceStart mainly provides progressive training plans.
How to reason through it
Write a plain-language expectation for each query before assigning priority. Inspect representative search results as contextual evidence, not as a definitive statement of every user’s intent. Check reviews, interviews and the product workflow where available. A query can have mixed intent; record that ambiguity instead of forcing it into a confident label.
The next practical move
Prioritise planning language when the product demonstrates a usable plan. If tracking exists only as a supporting feature, explain its role accurately. Avoid promising route discovery if the app has no route database. Connect the chosen intent to the opening screenshots so people can recognise the answer to their question immediately.
Swipe horizontally to compare the decision map →
Three meanings hiding behind a running keyword: illustrative decision map
Decision area
What to inspect
What it means for the work
Running routes
Find a suitable place to run
Needs route discovery evidence
Running tracker
Record activity and review performance
Needs actual tracking workflow
Running plan
Know which session to complete next
Fits a progressive training product
The trap to avoid
Do not equate shared vocabulary with shared intent. “Budget planner” may refer to personal spending, event planning or a business forecast. A useful cluster groups a similar expected outcome, not merely similar words.
When the situation gets more complicated
For mixed-intent phrases, decide whether a broader promise would improve comprehension or dilute it. Record the trade-off. You may choose a narrower phrase with a clearer fit rather than compete for an umbrella term that demands features you do not offer. Validate the opening message with people who resemble the intended audience, and ask what they expect to happen after installation.
FROM METHOD TO DELIVERABLE
An intent assessment with confidence and a challenge
Continue the fictional PaceStart running-app example. Confidence here concerns the interpretation, not a probability of ranking or conversion. A useful intent record includes the evidence that could overturn it.
Swipe horizontally to read the complete example →
Illustrative intent hypotheses. These are not measured distributions of searchers.
Query and interpretation
Confidence boundary
What would challenge the interpretation
Running plan / decide which session comes next
Plausible fit with a progressive training product; validate locally
Relevant users expect coaching features the app lacks
Running tracker / record a completed activity
Meaning may cover GPS, manual logs or device integrations
Research shows a required tracking mode is unsupported
Running routes / find somewhere to run
Route discovery seems likely but the use context is unresolved
Users mainly want to save familiar routes rather than discover new ones
Beginner running / ambiguous task
Audience cue is clearer than the desired workflow
Users split between plans, route ideas and injury-related advice
Use a short validation interview
Show a phrase without your proposed headline and ask what the person would expect an app to do. Ask them to describe the first task after opening it, the information they would need and what would disappoint them. Then show the listing and ask what they think it promises. Keep responses separate from your interpretation. A small qualitative review can reveal misunderstandings; it does not measure market-wide demand.
Handle mixed intent explicitly
When answers differ, record the alternatives instead of selecting the one that makes your product look best. Decide whether the listing can serve a coherent subset with an honest promise. For “beginner running,” a training-plan app can explain progressive sessions while avoiding claims about route discovery or medical advice. A broad query does not require a broad unsupported product promise.
Connect intent to an acceptance check
Before approving creative, ask whether a person expecting the selected job can identify the relevant product evidence. For a training plan, show a genuine next-session workflow rather than an unrelated activity chart. If the research suggests that the essential workflow is absent, revisit targeting. Better captions cannot solve a fundamental product mismatch.
A competitor ranks for all three phrases. Does that establish that PaceStart should target all three too?
Reveal the reasoning
No. The competitor may serve a broader product need or attract a different audience. Inspect what it actually offers. Competitor visibility can generate a research question, but your own product support determines whether the promise is defensible.
Your working brief
Use your own app for this exercise. Fill in the six fields below and attach the evidence behind each answer. If you cannot complete a field, write the missing question and the person who can resolve it. Do not replace unknowns with optimistic assumptions.
Exact query and market
Expected user action
Likely audience and context
Evidence supporting the interpretation
Ambiguous or competing meanings
Product fit and matching screenshot proof
For each field, distinguish an observation from an interpretation. Add the date and source where available, then identify the decision that follows. A colleague should be able to trace your recommendation back to the evidence without needing you to explain the entire project again.
Write a value proposition that answers the selected intent with product evidence.
The worked example is an original editorial framework. Its fictional app, decisions and scenarios are not an industry dataset or a claim of measured uplift.
