Anchored App Brief

The Anchored App Brief is a short, structured description of what your app actually is. Every recommendation engine reads it before it analyzes your listing, so the brief is the single place you correct the platform’s understanding of your app.

You’ll find it in the project’s Settings → Brand Identity tab, in the App Brief card.

Why it matters

Engines infer a lot about your app from its title, subtitle, description, and screenshots. When that inference is slightly off — a fitness app read as a nutrition app, a legitimate marketed claim read as an unsupported one — the recommendations drift with it. The brief pins down the facts the engines would otherwise guess at, so the analysis argues from what your app is rather than what its store text happens to imply.

Five engines read the brief on every audit: Store Text, Competitor Discovery, Intent, Sentiment, and Localization Analysis. It’s injected into each engine’s prompt as a shared context block.

Two concrete effects worth knowing:

Real value props stop false-claim flags. The Store Text engine flags claims in your listing that its factual-grounding pass can’t support. Anything you’ve listed as a real value prop is treated as an intentional, legitimate claim and is not flagged for removal.
Off-limits patterns keep recommendations in-category. Categories, intents, and keywords you mark off-limits are hard-stops. Competitor Discovery drops candidates that only match an off-limits intent, and the engines won’t recommend copy or keywords that pull your app toward an off-limits area.

The four fields

The brief has four fields. Keep each one factual and short — this is a reference the engines read, not marketing copy.

Positioning sentence — one sentence: what the app actually is.
Real value props — a list of the legitimate marketed claims your app makes. Engines treat these as intentional, not as claims to flag.
Intended user — one short paragraph: who installs the app and why.
Off-limits patterns — three comma-separated lists (Categories, Intents, Keywords) naming areas that are hard-stops for category-irrelevant recommendations.

Where the brief comes from

The brief is generated automatically the first time you run an audit on a project that doesn’t have one yet. The generator reads your live listing (title, subtitle, description, first screenshot) plus a summary of recent reviews and writes the four fields for you.

The footer under the card records its provenance:

Generated from your current listing — written by the generator, never edited.
Authored by you — you created or fully replaced it.
Generated … edited by you — generated, then edited by hand.

Once you edit any field, the brief is yours. The platform never overwrites your edits on a later audit.

Drift: when the brief no longer matches the listing

Every audit compares the listing it just analyzed against the listing the brief was generated from. When they’ve diverged enough, the audit marks the brief as drifted and an amber Brief drifted badge appears on the App Brief card (and a small marker on the project tile).

Drift is a signal, not a failure. The audit still runs against the brief you have — drift just tells you the brief may be describing an older version of your listing, so it’s worth a look.

There’s one case where the platform refreshes the brief for you: if the brief was machine-generated and you’ve never edited it, an audit that detects drift regenerates it from the current listing. When that happens you’ll see “Brief auto-updated on the last audit run from your current listing.” A brief you’ve edited by hand is never auto-refreshed — your version stands until you change it.

Reviewing and refreshing

Review it

Open Settings → Brand Identity → App Brief and read the four fields against what your app is today. This is the highest-leverage thing you can correct: every one of the five brief-aware engines inherits whatever is written here.

Refresh from the live listing

Use Refresh from listing to rebuild the brief from your current store content. This re-runs the generator against the live listing and replaces the fields. Use it after a listing rewrite when you want the machine’s read of the new content rather than editing by hand.

Edit by hand

Change any field and click Save. Your edits are locked in and the provenance flips to reflect your authorship. This is the right path when you know something the listing doesn’t say plainly — the true intended user, a claim that’s legitimate but reads as unsupported, a category the engines keep wandering into.

Verify it worked

After saving or refreshing, the App Brief card shows your current values and the footer reflects the new provenance and date. On the next audit, the brief-aware engines argue from the updated facts, and the Brief drifted badge clears once the brief matches the listing it was measured against.