How to Get More App Store Downloads in 2026: An ASO Playbook

An app with 60,000 App Store impressions a month and 1,400 installs is converting at 2.3%. Doubling that rate adds roughly 1,400 installs a month without a single extra impression, and because Apple’s ranking systems read installs as a relevance signal, those installs then earn you more impressions. That loop is where install growth compounds. The entry point to it sits in two assets most teams revisit least often: in search results, your icon and your first screenshot decide whether an impression becomes a tap at all.

This playbook walks the App Store surface as it behaves in 2026: which metadata fields are indexed and how heavily, how to mine Apple Search Ads for organic keyword decisions, how to run genuine A/B tests inside App Store Connect, how the rating prompt API is limited, and what App Review and the editorial team respond to. Google Play differences are flagged where they matter, since most teams ship both.

The conversion flywheel behind your install count

App Analytics in App Store Connect gives you a three-stage funnel. Impressions count every time your app appears in a search result, a browse shelf or an editorial collection. Product page views count the people who tapped through. Downloads are split by source type: App Store Search, App Store Browse, App Referrer, Web Referrer and Institutional Purchase.

Two paths lead to an install and they exercise different assets. A user can tap GET straight from the search results row, in which case your icon, name, subtitle, star rating and first screenshot did all the work. Or they open the product page, where the gallery and app preview take over. Search-heavy apps often convert a large share of installs with no page view at all, which is why the icon deserves more attention than the description.

Segment before you diagnose. Flat App Store Search impressions point to a keyword and ranking problem. Healthy impressions with weak page views mean your search-result presentation is failing. Healthy page views with weak downloads mean the gallery is failing. Web Referrer downloads dominating means store discovery is dormant and you are paying for every install off-store.

Apple has never published its ranking weights, so treat any formula with suspicion. The observable mechanism is directional: search results respond to text relevance from your indexed fields and to behaviour, including how often people who see you for a query tap and install, how the app is rated, and how much it gets used. That makes conversion the input. Improve the icon and the first two screenshots, more of the people already seeing you for “habit tracker” install, Apple reads that as evidence you answer the query, and shows you more often. The extra impressions convert at the new higher rate.

Sequencing follows from that. Fix presentation before expanding keyword coverage, because an app converting at 2% will bleed back ten new mid-volume rankings within weeks while an app at 5% holds them. If you are also running audience work across social channels, keep the loops separate in reporting; teams running their off-store push through LitFame’s growth services should tag those campaigns and read Web Referrer downloads on their own line rather than folding everything into one install total.

The name and subtitle carry different keyword weight

Three fields carry the bulk of Apple’s search index on iOS: the app name (30 characters), the subtitle (30 characters) and the keyword field (100 characters, hidden from users). Your developer name and your in-app purchase names are indexed too, which is worth remembering when you name a subscription tier. The description sits outside the index on iOS, as does promotional text. Keyword-rich description prose does nothing for App Store search, though it still affects conversion, and on Google Play the title, short description and full description all count toward ranking.

The name carries the heaviest keyword weight of any field, so a brand-only name spends your strongest asset on a term nobody searches. The standard pattern is a brand plus a descriptor separated by a colon or a dash: “Pocketlog: Habit Tracker” puts two real search terms in the most weighted field. Keep it readable. Apple rejects names that read as keyword lists, and a stuffed name converts worse anyway.

The subtitle is indexed at a weight generally understood to sit below the name and above the keyword field. It has a second job: it is the one line of copy visible in search results beneath your app name, so it must sell as well as rank. “Build habits with streaks & reminders” carries three indexable concepts and tells a browsing user what happens if they tap. Skip words already in the name, since repetition adds no index value and burns characters.

Both fields are per-localization and Apple builds its index per storefront language. An app localized into six App Store languages has six times the indexed metadata, often the cheapest impression increase available to a small team.

The keyword field, and what belongs in it

The 100-character keyword field is invisible to users and exists purely for the index. Use commas with no spaces after them, since a space costs a character and buys nothing. Skip any word already in your app name or subtitle, your own brand, your category name, plurals of words you already have, and connector words. Apple’s indexer generally handles singular and plural forms and combines your individual words into multi-word phrases, so “habit” and “tracker” entered separately can surface you for “habit tracker” without spending characters on the phrase.

Competitor brand names are a live risk. Apple has removed metadata and rejected updates over trademarked terms, and a rights holder can file a complaint that pulls your listing into a dispute. Small upside, large downside.

Think in coverage rather than individual keywords. A hundred characters holds roughly twelve to fifteen distinct words, and the goal is a vocabulary that combines into as many plausible queries as possible: the problem you solve, the object you act on, the outcome the user wants, the format, the audience. For a habit app that means streak, routine, daily, goal, ritual, discipline, planner, reminder. Each combines with words already in your name and subtitle.

Google Play has no keyword field. Rankings there come from the title, short description and long description, plus install and engagement behaviour, so the same research produces two different implementations.

Apple Search Ads as a keyword research instrument

The most reliable organic keyword data available to you comes from paid search, because it is the only place Apple reports real query behaviour on your own app. A Search Match campaign lets Apple choose which queries to show you on, and the Search Terms report then tells you exactly what people typed before tapping. A broad match campaign on a small seed list does the same with more control. Run either at a modest daily budget and treat the spend as research.

You get back real queries with impressions, taps, conversions and cost per install for your specific app. That last column is the useful one. A query where your ad converts at 15% is a query your product page already answers, so ranking organically for it will produce installs. A high-volume query converting at 2% is one you can rank for and gain nothing from. Rank ambition should follow conversion evidence.

The Search Ads keyword planner also shows a popularity score on a normalized scale for any term, the closest thing Apple publishes to search volume. Sort candidates into tiers: a few head terms you probably cannot win yet, a working set of mid-popularity terms where a well-converting app can realistically place, and long-tail terms that convert well and cost nothing to cover in the keyword field. Feed the winners into subtitle and keyword field revisions, then watch organic impressions for those queries over the following weeks.

The icon and the first two screenshots do most of the converting

In search results the App Store shows your icon, name, subtitle, rating and up to three portrait screenshots. On phone screens most users see the icon and one or two images before they scroll, and everything after that is read by a minority. Design accordingly.

Icons that perform tend to hold a single recognizable shape with high contrast against light and dark backgrounds, no text small enough to become a smudge, and a colour uncommon in the category. If every competitor in your results is blue, being orange is a measurable advantage before anyone reads a word. Test the icon at rendered size on a device rather than at 1024 pixels in a design file.

The first screenshot should state the single strongest reason to install, in a caption of five or six words above a clean device frame. The second shows the next reason. Captions do the persuading and the interface behind them provides evidence. A raw UI with no caption wastes the slot, because someone scrolling a results page cannot infer value from a list view they have never used.

App preview videos autoplay muted in results and on the product page. They help most where value is motion or interaction, such as games, editors and drawing tools, and least for utilities whose value is one static outcome. Make the first three seconds show the app doing the thing, since that is all most viewers watch. Apple requires previews to be captured from the app itself, so a stylized motion-graphics trailer gets rejected under the metadata rules.

Product page optimization and custom product pages

App Store Connect provides two tools that developers routinely confuse. Product Page Optimization is a genuine split test. You create up to three treatments against your original page, varying the icon, screenshots or app preview, and assign a percentage of App Store traffic to each. Apple reports impressions, conversion rate and improvement with a confidence indicator, and a test can run for up to 90 days. Icon treatments require the alternative icons to ship inside a build, so plan them one release ahead.

Run one variable at a time. Testing a new icon and a new screenshot set together tells you the combination won and nothing about which change caused it. Give a test enough traffic that the confidence indicator settles instead of calling it after a promising first week, and be ready for a result showing your challenger losing. That is still useful, and keeping the original costs nothing.

Custom product pages solve a different problem. You can create up to 35 alternative pages, each with its own screenshots, promotional text, app preview and URL. They never appear in App Store search. You point traffic at them: a TikTok campaign about meal planning lands on a page whose screenshots are all meal planning, while your Reddit link lands on a page about the privacy model. Conversion on well-matched custom pages typically runs well above the default page, because the promise in the ad and the promise on the store page are the same promise.

Custom pages only pay off if you have traffic to send them, which makes off-store audience building the real constraint for most indie teams. If you are building that audience alongside the store work, you can set up a LitFame account and run the social side against the same custom page URLs, since each custom product page reports its own impressions and conversion in App Analytics.

Ratings and reviews, handled legally

Star rating is visible at every stage of the funnel and moves conversion more than any line of copy. It also carries real legal exposure. In the United States, the FTC Rule on Consumer Reviews and Testimonials has been in force since October 2024. It prohibits creating, buying or selling fake consumer reviews, including reviews from people who do not exist and insider reviews that hide the relationship, and it prohibits compensation conditioned on the sentiment of a review. Violations can carry civil penalties per violation, with the maximum adjusted annually. Apple treats fabricated ratings and paid review schemes as discovery fraud under its App Review Guidelines and removes apps and developer accounts over it. Google Play prohibits incentivized ratings in its Developer Program Policies. Buying ratings risks your app, your account and your company at once.

The legitimate mechanism is the system rating prompt. On iOS you request it through StoreKit: recent SDKs expose a requestReview API on AppStore and a requestReview environment value in SwiftUI, while older code called SKStoreReviewController. You do not control whether the prompt appears. The system decides, and Apple caps display at three times per app per user in any 365-day period. Users can turn prompts off in Settings, and they are suppressed in TestFlight builds.

With so few chances, timing is the whole game. Trigger after a moment of demonstrated success: a completed workout, a finished export, a seventh consecutive day of use, a resolved support ticket. Avoid launch, onboarding, the moment after a paywall, and any button tap. Apple requires that the prompt not be attached to a user action, so a “Rate us” button must deep link to your App Store page with the write-review action appended to the URL.

Two more mechanics matter. Since iOS 11 ratings carry across versions by default, so a bad stretch follows you. App Store Connect lets you reset the summary rating, effective with your next version release: written reviews stay visible while the numeric average starts fresh. Use it only after genuinely fixing what people complained about, since a reset discards your good ratings too. And reply to reviews. Responses appear publicly, the user is notified, and reviewers often raise their rating after a real fix ships.

What App Review actually rejects

Rejections cost days, and a handful of guidelines account for most of them. Guideline 2.1, app completeness, catches builds with crashes, broken links, placeholder content, or a login wall with no working demo account in App Store Connect. Supply credentials for anything gated, social sign-in paths included, and spell out in the review notes how to reach features that are hard to find.

Guideline 2.3 covers accurate metadata, where ASO ambition collides with review. Screenshots must show the app in use rather than lifestyle imagery or claims the app does not deliver. Names and subtitles that read as keyword strings get rejected, as do descriptions of features requiring hardware you do not support.

Guideline 4.2, minimum functionality, hits thin apps: a repackaged website, a single static feature, a template build shared with dozens of others. Guideline 4.3 covers spam and duplicates, which is why shipping many near-identical apps from one account is a losing strategy in 2026. Guideline 3.1.1 requires in-app purchase for digital content consumed in the app, with the carve-outs Apple has published for reader apps and, in some jurisdictions, external purchase links under separate entitlements.

Privacy is the fastest-growing rejection category. Your App Privacy details must match what the app actually collects, third-party SDKs included. Apps using certain APIs need a privacy manifest declaring an approved reason, and SDKs on Apple’s list must ship their own manifests and signatures. Apps supporting account creation must offer in-app account deletion. All of it is checked.

How featuring actually happens

Editorial featuring on the Today tab and in curated collections is decided by Apple’s editorial teams, who work by region and category. There is no purchase path. There is a submission path: App Store Connect includes a featuring nomination form where you describe the app, the moment, and what is new. Submit several weeks ahead of the date you care about, since editorial calendars are planned well in advance and a nomination filed during launch week is almost always too late.

What editors respond to is fairly consistent. Adoption of new Apple platform capabilities is the strongest lever, because Apple has an interest in showcasing what its own releases enable: widgets, Live Activities, App Intents and Shortcuts, controls, watchOS and visionOS versions, new frameworks in the year they ship. Design quality matters, as does accessibility that works with VoiceOver and Dynamic Type. Localization into the storefront language makes regional featuring possible at all, and a genuine seasonal hook gives an editor a story to write.

Treat featuring as an accelerant on a business that already works, since a Today feature produces a large single-day spike and apps with poor retention convert that into a spike-shaped chart and nothing else. The teams who compound a feature have onboarding that holds, a rating prompt placed after a real success moment, and a steady channel of their own demand, whether that is search, email, or social audiences built through LitFame’s platform services. Nominate anyway. It costs an hour, and In-App Events, which surface timed content on your product page and in search, give you a second editorial surface on your own schedule.

Which levers to pull first

Effort and effect are unevenly matched across ASO work, and small teams lose months at the wrong end of this table. The ranges below describe what teams commonly observe rather than any published Apple figure, and your category will move them.

A sensible first quarter: week one, pull the App Analytics funnel and identify which stage is failing. Weeks two and three, run a Search Match campaign and collect query data. Week four, rewrite the subtitle and keyword field against that data and ship with your next version. Weeks four through ten, run a single-variable PPO test on the icon or the first screenshot. Throughout, move the rating prompt to a success moment and start replying to reviews.

Frequently Asked Questions

How many times can the iOS rating prompt appear for one user?

Apple limits the system rating prompt to a maximum of three displays per app, per user, in any 365-day period. You request it through the request-review API and the system decides whether to show it, so a request does not guarantee a prompt. Users can disable prompts entirely in Settings, and prompts are suppressed in TestFlight builds. Because your chances are so limited, request only after a clear success moment inside the app.

Does the App Store description affect keyword rankings?

On iOS the description is not part of Apple’s search index, so keywords there do not influence ranking. Only the app name, the subtitle and the hidden 100-character keyword field feed the index, along with in-app purchase names and your developer name. The description still affects conversion for users who read it. Google Play differs: title, short description and full description are all indexed, so the same research produces two implementations.

Can I offer users a discount for leaving a five-star review?

No. Under the FTC Rule on Consumer Reviews and Testimonials, in force since October 2024, offering compensation conditioned on the sentiment of a review is prohibited in the United States and can carry civil penalties. Apple treats manipulated ratings as discovery fraud and removes apps and developer accounts, and Google Play prohibits incentivized ratings in its Developer Program Policies. Ask for feedback at a genuine success moment and accept whatever rating follows.

What is the difference between custom product pages and product page optimization?

Product Page Optimization is an A/B test: up to three treatments run against your live page using App Store traffic, and Apple reports which converts better with a confidence indicator. Custom product pages are alternative pages with their own URLs that never appear in App Store search. You send external traffic to them so the creative in an ad matches the creative on the store page. One measures, the other targets.

Do my ratings reset when I release a new version?

Since iOS 11 ratings carry over across versions automatically. Resetting is a deliberate choice: in App Store Connect you can reset the summary rating, and it takes effect when your next version goes live. Written reviews remain visible while the numeric average starts from zero submissions. Use it only after fixing the underlying problem, because you also discard every positive rating and rebuild from a very small sample.

How do I get my app featured by Apple?

Submit the featuring nomination form in App Store Connect several weeks before the date you want, describing the app and what is new about it. Editors favour apps that adopt recent Apple platform capabilities such as widgets, App Intents, Live Activities or new device support, alongside strong design, working accessibility and localization for the storefront. Nothing is guaranteed and Apple gives no advance notice, so plan your launch without depending on it.

How long does an ASO change take to show results?

Metadata changes to the name, subtitle or keyword field ship with a version and typically take two to six weeks for rankings to settle, since Apple re-indexes and then needs behavioural data at the new position. Conversion changes such as screenshots show movement faster, often within two weeks of sufficient traffic. Rating improvements are slowest, because the displayed average moves only as new ratings accumulate against your existing volume.

Is Apple Search Ads worth running for a small app?

Even at a small daily budget it is the only source of real query data for your specific app, including which searches convert into installs and at what cost. A Search Match campaign run for two or three weeks produces a Search Terms report telling you which keywords deserve space in your subtitle and keyword field. Treat the spend as research first, and scale it as acquisition only where the cost per install works.