Yannick
Audit sequence
Diagnose visibility before changing creative
Verify the listing
Check metadata, assets, localizations, ratings, and release state.
Find the bottleneck
Separate discovery, conversion, trust, and retention problems.
Ship one hypothesis
Change one meaningful variable and define the expected result.
App name and subtitle match the strongest search intent.
Keyword field avoids repeats, protected terms, and irrelevant words.
Screenshots explain the product before the user reads the description.
Feature graphic, icon, and preview video match the same visual system.
Store listing is localized with translated screenshots where the market matters.
Ratings, reviews, and conversion data have a clear next action.
How to use this checklist
Freeze the store, country, language, listing version, and date first. Mark each check pass, fail, unknown, or not applicable; attach evidence; then score only the failed items. An unknown is not a pass—it is an instrumentation or ownership gap.
The 40-point ASO audit checklist
The order matters. Verify the live store state before reviewing copy, diagnose discovery before redesigning screenshots, and confirm the measurement plan before launching an experiment. This prevents a polished asset from hiding a metadata, localization, or product-quality problem.
Audit layer 1
Store foundation
- 1The live binary, listing version, price, countries, and release state match the audit scope.
- 2Support, privacy-policy, marketing, and legal URLs load on mobile without redirects or errors.
- 3Category, age rating, content declarations, and app availability match the product users receive.
- 4The app name, publisher identity, and brand are consistent across both stores and the website.
- 5One owner and one dated evidence source are recorded for every audit finding.
Audit layer 2
Apple metadata
- 6The 30-character app name states the product or strongest differentiator without repetition.
- 7The 30-character subtitle adds meaning instead of repeating words from the name.
- 8The localized keyword field stays within 100 bytes and avoids app names, company names, duplicates, and protected competitor terms.
- 9Promotional text, description, and What's New copy support conversion without making unverifiable claims.
- 10Relevant App Store tags are reviewed for the US storefront and irrelevant tags are deselected.
Audit layer 3
Google Play metadata
- 11The title and short description communicate the core job before truncation on a small screen.
- 12The full description uses natural, specific language and does not repeat keywords mechanically.
- 13The category, tags, contact details, and developer profile support the same positioning.
- 14Data safety, ads, target-audience, permissions, and content declarations match actual behavior.
- 15Search-term custom store listings are considered where Play Console shows a distinct qualified audience.
Audit layer 4
Discovery and keywords
- 16Branded, category, problem, feature, and competitor-observation terms are separated by intent.
- 17Every priority keyword has a fixed store, country, language, baseline rank, and decision owner.
- 18High-impression, low-conversion terms are reviewed for message mismatch before more metadata is added.
- 19Competitor movement and neighboring search results are saved beside rank changes.
- 20Release, metadata, pricing, localization, and creative changes are annotated on the ranking timeline.
Audit layer 5
Creative assets
- 21The icon is recognizable at search-result size and does not depend on small text.
- 22Screenshot one communicates the primary outcome in one glance.
- 23The first three screenshots form a coherent promise → proof → differentiator sequence.
- 24Captions remain readable on small phones and describe evidence visible in the product UI.
- 25Device frames, status bars, typography, and brand treatment are consistent across the set.
- 26iPhone, iPad, Android phone, and tablet assets use the right aspect ratios and current UI.
- 27The Google Play feature graphic works as a 1024×500 promotional asset without critical edge content.
- 28App previews or promo videos open with useful product footage, work muted, and match the listing story.
Audit layer 6
Localization
- 29Priority markets are chosen from demand, conversion, revenue, and product readiness rather than translation cost alone.
- 30Metadata and screenshot captions are translated for local intent, not copied word-for-word.
- 31Screenshots show the correct in-app locale, units, dates, currency, and regional content.
- 32Long text, CJK, RTL, and line breaks are visually reviewed in every exported size.
Audit layer 7
Trust and product quality
- 33Rating trends, review themes, replies, crash signals, and Android vitals have an owner and next action.
- 34The listing does not promise features, pricing, devices, or integrations the current build cannot deliver.
- 35Recent reviews and support issues are checked for objections the screenshots or copy should answer.
Audit layer 8
Measurement and experiments
- 36Search visibility, product-page views, conversion, activation, retention, and revenue are reported separately.
- 37Apple Product Page Optimization and Google Play store listing experiments each test one clear hypothesis.
- 38A baseline, minimum runtime, target market, primary metric, guardrail, and stopping rule exist before launch.
- 39Custom product pages or custom store listings map a defined audience to matching creative and copy.
- 40The audit ends with a ranked backlog, owner, due date, expected impact, and measurement plan.
Diagnose metadata before changing creative
Check app name, subtitle, short description, full description, and keywords first. Apple and Google both keep visible metadata tight: Apple app names and subtitles are limited, while Google Play titles and short descriptions have their own character limits. If the main value proposition is not clear in those fields, screenshots have to work too hard.
Audit the first impression as one system
Review the first three screenshots, icon, feature graphic, and video together. Google says preview assets help promote apps across Play surfaces, and Apple lets you show up to ten screenshots plus app previews. Your audit question is simple: would a cold user understand the app from the first screenful?
Audit localization market by market
A translated description with English screenshots is usually a half-finished localization. Check whether your screenshots, captions, metadata, and screenshots match the market. Google and Apple both support localized store assets, so high-potential markets deserve localized creative, not only translated strings.
Turn findings into testable hypotheses
Apple Product Page Optimization lets you test product-page treatments that use alternate icons, screenshots, and app previews against the original page. Google Play has store listing experiments. If an asset is important enough to debate, it is usually important enough to test with a clean hypothesis, one primary metric, and a defined stopping rule.
Score issues by impact, evidence, and effort
A useful audit ends with a ranked backlog, not a page of observations. Prioritize issues that affect the first impression or block trust, then improvements with measurable growth potential, and finally visual polish. Add the evidence behind each item so the team knows whether it came from store data, user feedback, or a design review.
P0
Blocks trust
Broken assets, misleading claims, or missing localization.
P1
Limits growth
Weak search coverage or unclear first screenshots.
P2
Improves polish
Consistency, spacing, and secondary-message refinements.
Use one scoring model for every finding
Score impact, evidence, effort, and confidence from 1 to 5. A practical priority score is (impact × evidence × confidence) ÷ effort. The formula does not replace judgment; it makes tradeoffs visible when metadata, creative, localization, and engineering findings compete for the same release.
| Field | 1 means | 5 means | Evidence to attach |
|---|---|---|---|
| Impact | Small polish gain | Blocks discovery, trust, or conversion | Affected market, traffic, and funnel stage |
| Evidence | Opinion or isolated comment | Repeated store, experiment, review, or product data | Source link, date range, and sample size |
| Confidence | Several plausible causes | Direct cause or controlled result | Alternative explanations checked |
| Effort | One copy or configuration edit | Multi-team product or release change | Owner, dependency, and estimate |
Official sources
- Apple App Store Connect app information reference
- Apple Product Page Optimization
- Google Play preview assets
- Google Play app discovery and ranking guidance
- Google Play store-performance analysis
Practical deep dive
Prioritize an ASO audit by bottleneck
An audit is useful only when it separates discovery, conversion, product quality, and measurement. Score the evidence, identify the narrowest bottleneck, and ship one hypothesis with a named owner. A long checklist without priorities tends to produce simultaneous changes that cannot be evaluated.
| Layer | Evidence to collect | Typical finding | Next action |
|---|---|---|---|
| Discovery | Queries, ranks, impressions, metadata by locale | Relevant terms are absent or mis-mapped | Revise one keyword cluster and field map |
| Conversion | Product-page views, installs, assets, traffic source | First frames do not answer the visitor intent | Test one message and proof sequence |
| Trust and quality | Ratings, reviews, crashes, support themes | Listing promise exceeds product experience | Fix the product or narrow the claim |
| Retention and revenue | Activation, trial, purchase, retention by cohort | Acquisition gains attract low-fit users | Use downstream quality as a guardrail |
Create an evidence sheet, not a screenshot collection
For every market, capture the current metadata, asset order, listing variants, rank and conversion baselines, release date, and major traffic sources. Add the review themes and product-quality signals that can explain why a listing promise succeeds or fails after install.
Label facts, interpretations, and hypotheses separately. A falling conversion rate is a fact; a weak first screenshot is an interpretation; changing the first-frame promise is a hypothesis. This distinction prevents the audit from presenting opinions as causes.
Use an impact-confidence-effort queue
Prioritize work that affects a large, relevant audience, has a plausible evidence chain, and can be measured cleanly. Correct policy, metadata, localization, or broken-asset problems before debating small visual preferences.
Record the owner, release, target metric, guardrails, and review date for every selected action. Re-audit after meaningful releases or when the performance baseline changes, not simply because a calendar reminder says to redesign the listing.
Write an audit decision record
Summarize the diagnosed bottleneck, the strongest supporting evidence, alternative explanations, and the action selected. Include the affected countries, devices, listing variants, traffic sources, and the data window. This turns the audit into a decision that can be reviewed rather than a collection of opinions.
Define a primary outcome and at least one quality guardrail. A screenshot test may target product-page conversion while monitoring activation or trial quality. A metadata update may target relevant query visibility while watching conversion so broader reach does not dilute visitor fit.
At the review date, record whether the result supported, contradicted, or failed to resolve the hypothesis. Keep inconclusive tests in the history. They prevent the same weak idea from being repeated and reveal when the real limitation is traffic volume or measurement quality.
Close the audit only when every selected action has an owner, release vehicle, verification step, and measurable follow-up. Move useful but lower-confidence ideas into a backlog rather than mixing them into the current test.
This decision record should be short enough to update after every meaningful listing release.
Frequently asked questions
What should an ASO audit include?
Review metadata, keywords, screenshots, icons, videos, localization, ratings, reviews, experiments, product quality, conversion, and downstream user quality.
How often should I audit my app listing?
Audit before a launch, major feature or pricing change, localization push, and whenever ranking, conversion, or user-quality movement requires a diagnosis.
Can I test ASO changes?
Yes. Apple offers Product Page Optimization and Google Play supports store listing experiments. Define a hypothesis, metric, audience, and guardrails before launch.
Sources and next steps
- Apple: Product Page Optimization
- Google Play: store listing experiments
- Google Play: store listing best practices
Continue with ASO 2026 guide, ASO keyword tracking guide, and screenshot A/B testing guide.
