Knowledge base
What we actually know about ASO and Apple Ads.
Expert: with sources and numbers. Plain: the same, without them, for beginners.
The reference we work from. Every fact carries a visible evidence tier and the date it was last checked against its source, including when that check is old.
117 pages
250 claims
220 documented
A Documented. Apple or Google states it in their own documentation, with an exact URL and a verification date. Every numeric limit, policy rule and metric definition is [A], or it is not stated as a fact. 220 claims.
B Established practice. Not documented by the platform, but consistently reported by two or more independent credible practitioners. Usable for planning, always labelled. 22 claims.
C In circulation. Single-source or unevidenced. Recorded so you know it exists and know not to rely on it. Never stated as fact. 8 claims.
Foundations
- How app discovery works The three ways people arrive at an app (search, browse and outside link) and why the split decides where the work goes. 2026-09-20
- App Store surfaces (iOS) Every place an iOS app can appear, what each one draws on, and which of them are new enough that most material predates them. 2026-09-20
- Google Play surfaces Where a Play app can appear, and why Google describes its surfaces far less precisely than Apple does. 2026-09-20
- Ranking inputs, what is known and what is assumed The complete list of what Apple and Google actually state about ranking. And the boundary where everything else begins. 2026-09-20
- How the two stores differ structurally "The differences that change the work: indexed fields, release binding, how much each store tells you." 2026-09-20
- Why almost everything is done twice The multiplication that makes ASO an operational problem rather than a writing problem, and how to keep it tractable. 2026-09-18
- Top charts, and what moves them What the stores document about chart position (almost nothing), what chart traffic is worth and why buying it is a bad trade. 2026-09-20
- AI as a discovery layer What each store has shipped, how an AI system assembles a recommendation, and how little of it a developer controls. 2026-09-22
App Store metadata (iOS)
- App Store metadata, overview and field map Every App Store text field, its limit, whether search indexes it, and whether changing it needs a new app version. 2026-09-19
- App name and subtitle "The two highest-weight text fields on iOS: their limits, Apple's rules and why truncation is a range, not a number." 2026-09-18
- The keyword field The hidden 100-byte field. Apple's own rules for it, the bytes-versus-characters question, and how to spend it. 2026-09-20
- Promotional text "170 characters, changeable without a release, and explicitly not a ranking input: the one iOS field you can move at will." 2026-09-18
- Description and What’s New 4,000 characters each, both version-bound, and neither named by Apple as a search input, which changes what they are for. 2026-09-18
- Indexed surfaces beyond the core fields Developer name, in-app purchase names and in-app event metadata are searchable text you control, bound to no app version. 2026-09-20
- Locales and storefronts 50 localizations, each storefront with default and additional languages, and the fallback chain that decides what users see. 2026-09-18
- Categories, age rating and privacy labels Three non-text properties that shape discovery and compliance, one of which is an indexed ranking input. 2026-09-20
- What requires a new app version Which metadata you can change on a live app today and which waits for the next release. Check before promising a date. 2026-10-06
- Brand and trademark in metadata What Apple's rules actually prohibit in names, subtitles and keywords. And where the risk sits when they are broken. 2026-09-18
- Localization coverage by storefront Which languages are indexed in which App Store storefronts, the lookup table behind every locale decision. 2026-09-20
Google Play metadata
- Google Play metadata, overview and field map Every Play store listing field, its limit, what Google says feeds search, and what is and is not documented. 2026-09-19
- App title 30 characters, indexed, visible everywhere, and governed by a policy that bans more than people expect. 2026-09-19
- Short description 80 indexed characters that most users read and most teams under-write. Play's highest-leverage field per character. 2026-09-18
- Full description 4,000 indexed characters, and the field where Google states most directly that keyword stuffing is against policy. 2026-09-18
- Developer name and package name One is metadata under policy; the other is permanent and the most consequential irreversible decision on Play. 2026-09-18
- Categories and tags A category and up to five tags. Both changeable without a release, and the closest thing Play has to structured keywords. 2026-09-20
- What changes without a release Play decouples the listing from the app release. But decoupled is not instant, and review still applies. 2026-09-18
- Writing for Play versus iOS Why the same copy cannot serve both stores. Not because of character counts, but because the indexed text does opposite jobs. 2026-09-19
Keywords
- Where search term data comes from The five sources of real query data, what each actually measures, and why none of them is a search volume report. 2026-09-19
- Third-party volume and difficulty scores What vendor keyword scores are actually made of, where they break, and how to use them without believing them. 2026-09-18
- Head versus long tail, and how to select Why a ranked list of terms is not a strategy, and the four tests a term has to pass before it earns space. 2026-09-20
- How the stores combine words Why you supply words rather than phrases, what is documented about combination, and what is inference. 2026-10-06
- Localisation as transcreation Why translating metadata produces text that ranks for nothing, and what to do instead. 2026-09-19
- CJK and RTL specifics Where the standard keyword rules stop working: the bytes question, word boundaries, and script mixing. 2026-09-20
- Brand, generic and competitor terms Three kinds of term with three risk profiles, and the line between metadata and ads, which is not where people assume. 2026-09-19
- Rank tracking and what it is worth What a tracked rank actually is, why it is a modelled sample rather than a measurement, and what to use it for. 2026-09-20
- Validating a term with paid before committing it Buying a term to find out whether it converts, before spending scarce metadata space and a release cycle on it. 2026-09-19
Creative and conversion
- App icon The one asset that appears on every surface, in every result, at a size where almost nothing is legible. 2026-09-19
- Screenshots (iOS) Current required sizes, how the scaling fallback works, and why the first three are a search asset rather than a page asset. 2026-10-06
- Graphics and screenshots (Google Play) Exact specs per asset and device type, and the device-fragmentation problem that has no iOS equivalent. 2026-09-20
- App previews and videos Specs on both platforms, the poster frame nobody sets, and why Apple's content rule limits what a preview can be. 2026-10-06
- Creative assets and the Asset Library Both stores now separate creative from the app release. Apple's Asset Library and new visuals went live on 2026-10-05. 2026-10-11
- The first impression frame What is actually visible before a user decides, and why most product page work is spent below that line. 2026-09-19
- Creative localisation The most commonly skipped lever in ASO, why it is skipped, and the cheap version that captures most of the value. 2026-09-18
- Ratings and reviews The prompt APIs and their limits, responding, the rating reset, and what rating actually does to discovery. 2026-10-06
- Review policy and rating crises What both platforms prohibit around ratings, where the line falls and what to do when a rating collapses for outside reasons. 2026-09-19
- Briefing and producing creative The division of labour between ASO and design, a brief template, and every store creative constraint in one table. 2026-09-19
Testing
- iOS product page optimization Apple's native A/B test: three treatments, 90 days, icon and screenshots and previews only. 2026-09-20
- Google Play store listing experiments Two variants, explicit confidence and MDE settings, text testing that iOS does not allow. And a metric that changed in 2026. 2026-09-18
- Designing a test "Hypothesis, one variable, a stopping rule decided in advance: the three things that separate a test from a watched change." 2026-09-18
- Reading results honestly "What a confidence interval says, why \"no difference\" is a real result, and the metric change that breaks comparisons." 2026-09-19
- Seasonality and confounds What a controlled test protects you from, what it does not, and the confounds that survive randomisation. 2026-09-19
- What native testing cannot do The gaps in both tools, the alternatives, and what each alternative actually measures. 2026-09-19
Custom pages and targeting
- iOS custom product pages Up to 70 alternative product pages, each with its own permanent URL. And, less widely known, its own keywords. 2026-09-20
- Google Play custom store listings Up to 50 listings with richer targeting than iOS, one constraint that shapes everything, and which listing a user sees. 2026-09-20
- Deep links, universal links and App Links What decides whether a click lands where it promised, and the Android capability that updates without a release. 2026-09-18
- Pre-registration and pre-orders "Building an install base before launch: a hard 90-day window on Play and an iOS rule that limits later changes." 2026-09-19
- Measuring custom pages What each platform reports per page, the thresholds that hide small pages entirely, and why a CPP comparison is not a test. 2026-09-19
Events, promotions and featuring
- iOS in-app events "Indexed, timed metadata configured outside the app version: every field, every ceiling and the scheduling rules." 2026-09-20
- Google Play promotional content Play's equivalent of in-app events, with an eligibility gate, a quarterly quota and a 60-day submission window. 2026-09-18
- LiveOps on Google Play What LiveOps actually refers to on Play, and why the name causes more confusion than the feature. 2026-09-18
- Featuring on the App Store The nomination mechanism, the lead times Apple states, and an honest account of what is and is not within your control. 2026-09-20
- Featuring on Google Play What Google documents about editorial selection, which is very little, and what follows from that. 2026-09-18
- Measuring events and featuring What each platform reports, and why the honest measurement of a feature is harder than the honest measurement of an event. 2026-09-18
- Seasonality and the ASO calendar Turning a demand calendar into a plan, given that every store lever has a different lead time. 2026-09-19
- Subscription offers and store promotional tools Offers that surface on the store rather than only inside the app, what each is for, and the limits. 2026-09-19
Analytics and measurement
- App Store Connect App Analytics Apple's own metric definitions, including a conversion rate that is not what most people compute. 2026-09-19
- Play Console acquisition reporting Google's metric definitions, the retention metric that needs no app open, and the 2026 change that broke old comparisons. 2026-09-20
- Where the two platforms are not comparable The page to hand someone who asks why iOS and Android conversion rates differ by a factor of three. 2026-09-18
- Data pitfalls "Unique versus total counting, aggregation, latency, thresholds, opt-in: failures that make a number wrong yet look right." 2026-09-18
- Benchmarks and attributing change What a published benchmark is actually worth, and how to claim credit for a change without overclaiming. 2026-09-19
Apple Ads
- Account structure and placements The four placements, what each is for, & where budget & keywords actually live. 2026-10-11
- Basic versus Advanced Two products, not two tiers: different pricing models, different ceilings, different control. 2026-09-18
- Apple’s recommended campaign structure The four-campaign structure Apple publishes, the reasoning behind it, and where it stops being useful. 2026-09-20
- Keywords and match types Broad and exact, which one you get if you do nothing, and the rule that separates paid keyword targeting from metadata. 2026-09-20
- Search Match and negative keywords Automated matching driven by your product page metadata, and the exclusion mechanism that keeps it useful. 2026-09-20
- Bidding, budgets and caps What you actually bid for, how to derive the number, the two bidding modes, and what the daily budget actually caps. 2026-09-20
- Audiences and targeting Every refinement available. And the one number that should stop you using half of them. 2026-09-18
- Creative and custom product pages in ads How ads are built from product page assets, how custom product pages match campaign to landing page, and the default ad. 2026-10-11
- Metrics and the order to optimise them Every Apple Ads metric with Apple's own definition, and the sequence that stops you fixing the wrong thing. 2026-09-18
- The Search Terms report as a research source The only real App Store query data a developer can get, and how to turn it into organic metadata. 2026-09-18
- Attribution AdServices, AdAttributionKit and SKAdNetwork, what each is for, and why Apple Ads is measured differently. 2026-09-18
- Halo, cannibalisation and incrementality The three arguments you will be asked to make about paid spend, and how far each can honestly be taken. 2026-09-20
- API and automation What is available programmatically, what is worth automating, and what should stay manual. 2026-09-18
Google Ads app campaigns
- Structure and assets You supply assets, not targeting, the structural inversion that makes App campaigns unlike Apple Ads. 2026-09-18
- Bidding and signals What you control (an objective, a target, a budget) and the rule that decides whether the campaign can work at all. 2026-09-18
- How App campaigns interact with Play custom store listings The page where Play ASO and Google UA actually meet. And the one topic in this section worth full attention. 2026-09-20
- Measurement What App campaigns report, what Play reports, and why the two will not agree. 2026-09-18
Wider growth context
- Attribution fundamentals What attribution is claiming, the choices baked into every system, and why numbers never reconcile. 2026-09-19
- Mobile measurement partners What an MMP does that the platforms do not, and what an ASO practitioner should ask of one. 2026-09-18
- SKAdNetwork and AdAttributionKit Apple's privacy-preserving attribution frameworks, what they deliberately hide and what an ASO practitioner needs from them. 2026-09-18
- Play install referrer Android's first-party install attribution signal, simpler than iOS's, and with one behaviour worth knowing. 2026-09-18
- How paid and organic interact The mechanism connecting them, which direction each effect runs, and how much of this is actually documented. 2026-09-18
- Web to app Sending web traffic into the store or the app, and the ASO-relevant parts of it. 2026-09-18
Policy and process
- App Review Guidelines for metadata Every metadata-relevant guideline in one place, with what each actually prohibits. 2026-09-18
- Google Play policy for metadata The Metadata policy and its neighbours: what is prohibited in titles, icons, developer names, descriptions and graphics. 2026-09-18
- Common rejection reasons The metadata-driven rejections an ASO practitioner causes or can prevent, and how to avoid each. 2026-09-18
- Release process. App Store What is bound to a release, what is not, and the sequence that decides when your metadata reaches users. 2026-09-21
- Release process. Google Play Listing changes decoupled from the app release. And the review, restart and publishing rules that decide when they land. 2026-09-18
- Roles and permissions What access an ASO practitioner actually needs on each platform, and the permissions that quietly block the work. 2026-09-18
- Trademark and brand in practice Two separate systems, platform policy and trademark law, and where the line between ads and metadata actually falls. 2026-09-18
- Managing metadata at scale Where manual metadata management breaks, what the APIs cover, and what should stay manual on purpose. 2026-09-19
- ASO team structures and roles The division of labour ASO requires, the three common team shapes and which dependencies block work when a role is missing. 2026-09-19
- Android vitals and technical quality The one part of Play discoverability owned by engineering, with the thresholds written down and the dates coming. 2026-09-19
Conflicts and corrections
Runbooks
- Analysing a competitor A repeatable teardown that produces decisions rather than a slide of observations. 2026-09-19
- Creating an in-app event end to end From idea to live, with the scheduling rules that decide the calendar. 2026-09-18
- Creating and managing custom pages and listings Building alternative store pages on both platforms, and keeping identifiers traceable afterwards. 2026-09-18
- Designing, running and reading a product page test One procedure, two platforms, with the differences called out where they bite. 2026-09-18
- Handling a rejection What to do in the first hour, how to decide between fixing and appealing, and how to stop it recurring. 2026-09-18
- Investigating a sudden drop or spike in installs A diagnostic order that eliminates the cheap explanations before the expensive ones. 2026-09-18
- Launching an Apple Ads discovery campaign and harvesting terms back into metadata Buying real query data on a platform that publishes none, and turning it into organic metadata. 2026-09-19
- Launching or optimising a new market Entering a storefront properly, or fixing one that was entered badly. 2026-09-18
- Producing a monthly ASO report that survives scrutiny A report that holds up when someone hostile reads it, which means stating definitions and separating what you can attribute. 2026-09-18
- Quarterly keyword refresh Reviewing and shipping the indexed text fields on a release cycle, without breaking what already ranks. 2026-09-19
- Submitting a featuring nomination The documented process, the lead times, and an honest statement of what it can and cannot achieve. 2026-09-18
Vocabulary
Questions people ask
No matches.