ASO Conversion Optimization

Getting found is aso-keyword-research/aso-metadata's job. This skill is what happens after someone sees the listing: does it convert a view into an install. aso has no visual or analytics data, this is qualitative guidance plus what the CLI can show you about the current listing and competitors.

Preflight: see the aso router skill.

Iron law

Conversion assets carry no searchable text. Icon, screenshots, preview video, and Custom Product Page content don't get indexed, so never claim screenshot captions or icon text rank for keywords; treat every recommendation here as a conversion lever only. Conversion rate and ratings are widely believed to feed back into ranking indirectly over time (see aso), which is why this work still matters for ASO, just not through indexing.

What the CLI can tell you

aso app <appId> --lang en-US        # your own listing: screenshot count context via description, ratings, version
aso search "<core keyword>" --limit 20 -c US   # top competitors for the terms that matter
aso app <competitorId> --lang en-US            # pull a competitor's listing to compare against

aso doesn't fetch actual screenshot images or icon files, so ask the user to share screenshots/icon directly (a screen share, a file, or a description) for real feedback; use the commands above to identify who to compare against and to check basic facts like current rating and version recency.

Icon

• No more than 2 visual elements. A cluttered icon reads as noise at thumbnail size.
• No text in the icon. It's unreadable at the sizes it actually displays (Spotlight, notifications, Settings), and Apple discourages it.
• Test legibility at actual thumbnail size, not the 1024×1024 master: view it in a search results mock-up or shrink it to ~60px before judging.
• Distinctive against the category's visual norms: if every competitor uses a blue gradient with a similar icon, sameness is not safety, it's invisibility.

Screenshots

• The first 2-3 screenshots carry most of the conversion weight. Most viewers never scroll past them; front-load the strongest material.
• Sequence: hook (first screenshot, the single reason to care) → value (what the app does for the user) → features (specifics, differentiation) → trust (ratings, awards, press, social proof).
• Headlines lead with an action verb and are specific, not generic: "Track every workout automatically" beats "Powerful Fitness Features." A generic headline is a wasted screenshot regardless of the visual behind it.
• Show the actual product UI, not just marketing illustration: users are judging whether the real app looks usable.
• Localize screenshot text per locale you're filling (see aso-localization), a screenshot with untranslated English text in a non-English storefront undercuts trust.

Preview video

• The value proposition must be clear in the first 3 seconds. Autoplay is brief and many viewers won't unmute or watch further.
• Must work with sound off: rely on on-screen text/captions and visuals, not narration, to carry the message.
• Show the real app in use, not an abstract brand film.

Custom Product Pages (CPPs)

• Up to 70 per app, each with its own unique URL.
• Each CPP requires App Review before it goes live, same as a metadata update, plan lead time.
• A CPP is not an A/B test by itself. It's a targeted variant you drive specific traffic to (a specific ad campaign, a specific channel, a specific audience segment); it only becomes a comparison when that traffic is externally split (e.g. across ad campaigns pointing to different CPP URLs). Don't present "we made a CPP" as if it already proves anything about conversion lift.

Product Page Optimization (PPO)

• Up to 3 treatments tested against the default page, for up to 90 days per test.
• Only icon, screenshots, and app previews (video) are testable in a PPO treatment: no title/subtitle/description/keywords changes inside a PPO test.
• This is Apple's native, traffic-split A/B test (unlike a CPP): it randomly assigns real App Store visitors, which is what makes its result a measured comparison, not a guess.
• One variable focus per treatment where possible (e.g. just the first screenshot) so a lift is attributable; a treatment that changes icon and all screenshots is hard to explain afterward even if it wins.

In-App Events

• Up to 10 concurrent events per app, each running up to 31 days.
• Events can surface directly in App Store search results and on the Today/Games/Apps tabs, an additional discovery surface, not just an in-listing banner.
• Use for real, time-bound moments: a seasonal challenge, a content drop, a live tournament, a limited sale, not evergreen features dressed up as an "event."
• Pair with aso-metadata's promotional text and What's New guidance so the messaging is consistent across every surface for the event's duration.

Workflow

1. Ask for the current icon and screenshots (or view them via aso app's listing link) and 1-2 competitor app IDs.
2. Pull competitor listings (aso app <competitorId>) and compare sequencing, headline style, and how recently their screenshots/preview look updated.
3. Give specific, per-asset feedback against the checklists above, not generic praise.
4. If the user has traffic to split (ad campaigns, multiple channels), recommend PPO for icon/screenshot testing and CPPs for channel-specific variants; if they don't have split-able traffic, say plainly that a CPP alone won't produce a measured comparison.
5. Hand off timely messaging (promotional text, What's New, In-App Events) to aso-metadata for the exact field content and character limits.

Red flags

• Claiming any of this affects search ranking. It doesn't; aso-metadata/aso-keyword-research own ranking.
• Treating a single Custom Product Page as proof of a conversion lift with no external traffic split.
• Recommending a PPO treatment that changes more than a couple of variables at once.
• Generic screenshot headlines ("Amazing Features") presented as if they were benefit statements.