The Metadata Score: How Sniffy Grades Your App's ASO Health

A score, not a verdict

Sniffy assigns every app a 0–100 metadata score across six weighted components. It is a diagnostic, not a verdict: a 63/100 tells you where your ASO health is weakest, not that your app is 63% as good as it could be. The point of the breakdown is to make it obvious which lever to pull first.

The six components and what they actually measure

• title · 20%

  Penalizes unused character budget, missing category keyword, or brand-only titles. A "Pawprint Habits" title scores well on brand recall but loses points for carrying no category keyword — Apple weights titles more than any other field.

• subtitle · 15%

  Looks at whether the subtitle carries a high-intent keyword as an exact phrase, whether it overlaps with title tokens (waste), and whether it stays under the 30-char cap.

• keywords · 20%

  Penalizes redundancy (tokens already in title/subtitle), format errors (spaces after commas burn bytes), plural forms (Apple stems them automatically), and forbidden terms (your own app name, your category, competitor brands per App Store Review §5.2).

• screenshots · 10%

  A description-density proxy. Apple's semantic search reads text rendered inside the first three screenshots, but Sniffy does not extract caption text directly — this score is an indirect signal you should verify by hand.

• ratingsAndReviews · 15%

  Average star rating and review volume. Apple weights both into ranking; a high-rating app surfaces higher on marginal keyword matches. The weight is intentionally bounded so a low-volume app with great metadata can still post a competitive score.

• keywordRankings · 20%

  A direct readout of your current rank distribution: how many keywords sit in 1–10, 11–30, 31–50, 51+, or are unranked. Tied with title as the heaviest weight — outcomes matter more than form.

How to read the breakdown

The overall number is the headline; the per-component sub-scores are the actionable part. A 63/100 that splits into title 70 / subtitle 55 / keywords 48 / screenshots 72 / ratings 80 / rankings 60 is telling you three things at once: your brand-side metadata is fine, your discovery-side fields (subtitle and keywords) are leaking the most weight, and your social proof is doing real work.

Pair the score with the report's top-three recommendations[] — those are ranked by impact-over-effort, and they typically attack the two lowest-scoring components first.

What the score deliberately doesn't include

Description is not scored on iOS — Apple does not index it for search, so the score weights it at zero (it shows up under screenshots only as conversion context). Localization is not scored in the overall — we surface a separate localization gap analysis instead, because mixing locales into a single number obscures which storefront actually needs work. Promotional text is not scored — it's a refresh channel, not a ranking channel.

How Sniffy measures this

The sample report shows the full metadata score block, including per-component notes that explain why each sub-score landed where it did. The weights object is returned in every report response so SDK consumers can build their own breakdown views.