// the ai discovery practice

The stores rank apps.
Machines now recommend them.

Apple's AI writes tags for your app. Siri surfaces it — or doesn't. Gemini builds Play collections around it — or around your competitor. And when users ask ChatGPT "what's the best app for X," someone gets named. We make sure it's you.

Get your recommendation auditSee AI Discovery Sprint
01
what changed

App store optimization was built for one behaviour: a human typing into a search box. That behaviour now shares the funnel with three others.

Assistants answer install-intent questions directly. Apple Intelligence surfaces apps by inferred capability. Play's generative collections assemble shortlists no keyword tool can see.

The listing is no longer just a search asset. It is training data — the tags, metadata, reviews, and press coverage that teach every machine what your app is for. AI Discovery is the discipline of authoring that understanding deliberately.

02
the discovery stack

Six workstreams. One recommendation surface.

01
RECOMMENDATION AUDIT
Where your app appears — and doesn't — across Siri suggestions, Play AI collections, and assistant answers for your category's install-intent prompts. Baseline share.
02
AI-TAG & METADATA ENGINEERING
Metadata structured for Apple's AI-generated tags and Play's semantic matching: capability-clear descriptions, entity-clean naming, feature phrasing machines can lift.
03
CAPABILITY EVIDENCE SYSTEM
The proof layer machines read: review themes, press coverage, listicles, and web presence aligned to the capabilities you want to be recommended for.
04
CUSTOM PRODUCT PAGE MATRIX
CPPs mapped to intent clusters — every paid channel, and every AI-referred visitor, lands on the page built for their question.
05
ASSISTANT ANSWER OPTIMIZATION
The app-recommendation twin of GEO: structured comparisons, category pages, and third-party validation in the sources assistants cite when naming apps.
06
SHARE-OF-RECOMMENDATION TRACKING
Monthly tracked-prompt panels ('best X app for Y') across ChatGPT, Perplexity, Gemini + store surface monitoring. Citation share, sentiment, position — beside your keyword ranks.
// signature metric

Share of
Recommendation.

Keyword rank tells you where you stand in a search box. Share of Recommendation tells you how often the machine's answer is your app.

We report both, monthly — tracked-prompt panels across ChatGPT, Perplexity, Gemini, Copilot; plus Apple Intelligence and Play surface monitoring. Citation share. Sentiment. Position.

// methodology preview lives at /data
// faq

Questions we get.

Does this replace ASO?+
No — it's built on it. ASO wins the search box. AI Discovery wins the recommendation. You need both, and the disciplines share the same underlying asset: your listing.
Can you guarantee Siri placement?+
Nobody can. We engineer the inputs Apple's models read — metadata clarity, capability evidence, review sentiment — and we measure how often you get surfaced. No fantasies.
Which assistants do you track?+
ChatGPT, Gemini, Perplexity, Copilot, plus store-surface AI features (Apple Intelligence app suggestions, Play's Gemini-picked collections).
Timeline?+
Metadata and tag effects land in weeks. Recommendation share compounds quarterly as review themes, press coverage, and third-party citations catch up to your positioning.
Games too?+
Yes — different surfaces (Play's game collections, App Store editorial), same discipline: engineer the inputs, measure the outputs.
Pricing?+
Practice add-on to any retainer, or the AI Discovery Sprint — a 4-week engagement that ships the audit, metadata engineering, and CPP matrix as a fixed package.
// ready?

Get your recommendation audit.

Where does your app appear when users ask machines what to install? We'll baseline your Share of Recommendation and ship the fix roadmap.

Book the audit →