/store-growth-audit

Audit Checklist P0–P4 — Foundations (25 items)

Stanza format, flags, status vocabulary, and new-app: semantics are defined in SKILL.md. Evidence keys (EV.*) are defined in detection-playbook.md. IDs are stable — never renumber.

P0 — Day-one money toggles

P0.1 Small Business Program enrollment — core

• detect: MANUAL Q1
• rule: ✅ enrolled · 🔴 eligible but not enrolled · 🟠 not sure · applies-if: <$1M proceeds/yr (else ⚪ ineligible)
• new-app: plan — apply as soon as the account has (or will have) paid transactions; 15% vs 30% commission
• fix: growth/indie-business → ASC manual (enroll under Agreements)

P0.2 Billing Grace Period + Billing Retry enabled — core

• detect: MANUAL Q2 (+ EV.subs to establish applies-if)
• rule: ✅ both enabled · 🟠 one, or unsure · 🔴 neither · applies-if: has auto-renewable subscriptions
• new-app: plan — enable at subscription setup; recovers 5–15% of revenue lost to failed cards
• fix: generators/subscription-lifecycle → /apple:subscription

P0.3 Analytics baseline + peer-group benchmarks

• detect: HYBRID — EV.analytics (report configured, baseline recorded) + MANUAL Q3 (benchmarks checked)
• rule: ✅ baseline snapshot recorded AND benchmarks checked this quarter · 🟠 one of the two · 🔴 neither
• new-app: defer — activates once the app has 30 days of store data
• fix: growth/analytics-interpretation → /apple:learn-from-store

P0.4 Keyword research done (autocomplete + competitors + Apple Ads discovery probe)

• detect: HYBRID — keyword research artifacts (.planning/ASO.md keyword table or equivalent) + MANUAL Q4
• rule: ✅ researched list incl. popularity data exists · 🟠 informal/partial research · 🔴 keywords guessed
• new-app: plan — research before first metadata, not after
• fix: app-store/keyword-optimizer + app-store/apple-search-ads → /apple:metadata (or /apple:aso if installed)

P1 — On-metadata ASO

P1.1 Title, subtitle, keyword field optimized — core

• detect: MCP EV.meta
• rule: ✅ all three fields near their 30/30/100 limits, no cross-field duplicate words, keyword field comma-separated without spaces · 🟠 fields present but wasteful (duplicates, spaces, unused chars) · 🔴 subtitle or keyword field empty/default
• new-app: plan — score the drafted fields in ASO.md by the same rule
• fix: app-store/keyword-optimizer → /apple:metadata

P1.2 Cross-localization indexing exploited

• detect: MCP EV.locales + EV.meta
• rule: ✅ extra indexed locales for the primary storefront carry distinct keywords (e.g. en-GB, es-MX for US) · 🟠 extra locales exist but duplicate the primary keywords · 🔴 primary locale only
• new-app: plan
• fix: app-store/keyword-optimizer (advanced-tactics §1) → /apple:localize

P1.3 Developer name is a deliberate, descriptive choice

• detect: MANUAL Q5
• rule: ✅ descriptive or deliberately branded · 🟠 never considered · 🔴 misleading/squatted (rare)
• new-app: plan — one-time strategic choice; heavy to change later
• fix: app-store/keyword-optimizer (advanced-tactics §12)

P1.4 Promoted in-app purchases with keyword-bearing display names

• detect: MCP EV.iaps (display names; promoted flag may need MANUAL confirm)
• rule: ✅ promoted IAPs configured with descriptive 30-char display names · 🟠 IAPs exist, none promoted or names generic ("Pro Upgrade") · 🔴 none configured · applies-if: has IAPs/subscriptions
• new-app: plan — up to 20 promoted IAPs, each an indexed search result card
• fix: generators/promoted-iap → /apple:iap

P1.5 In-app event metadata used as indexed surface

• detect: MCP EV.events
• rule: ✅ at least one event in the last 6 months with keyword-conscious name/description · 🟠 events exist but metadata is throwaway · 🔴 never used · applies-if: app has event-worthy moments (judgment — else ⚪)
• new-app: defer — activates post-launch
• fix: generators/in-app-events → /apple:event

P1.6 Description + promotional text working

• detect: MCP EV.meta
• rule: ✅ benefit-led first paragraph AND promo text in use (it updates instantly, no review) · 🟠 description fine but promo text empty · 🔴 feature-list-only description
• new-app: plan
• fix: app-store/app-description-writer → /apple:metadata

P1.7 Metadata written for Apple's AI tagging + App Store Tags curated

• detect: MCP heuristic on EV.meta (copy unambiguous about what the app does and who it's for?) + MANUAL ("App Store Tags reviewed — irrelevant tags deselected?")
• rule: ✅ literal, tag-friendly copy incl. screenshot captions AND tags curated in ASC · 🟠 clever/metaphorical copy or tags never reviewed · 🔴 misleading copy or wrong tags left live
• new-app: plan
• fix: app-store/keyword-optimizer (advanced-tactics §13 + §15) → /apple:metadata

P2 — Conversion assets

P2.1 Icon is deliberate and has been tested

• detect: HYBRID — EV.ppo (any icon experiment ever) + MANUAL
• rule: ✅ icon tested via PPO or consciously chosen against competitors · 🟠 default/first-draft icon never revisited · 🔴 placeholder
• new-app: plan — icon is the single biggest conversion lever in search results
• fix: generators/product-page-optimization → /apple:icon + /apple:experiment

P2.2 Screenshots: first three carry benefit captions — core

• detect: HYBRID — EV.meta where screenshot info is exposed; else MANUAL
• rule: ✅ first 3 screenshots lead with benefit captions readable at thumbnail size · 🟠 screenshots fine but captions absent/feature-speak · 🔴 raw UI dumps
• new-app: plan
• fix: app-store/screenshot-planner → /apple:screenshots

P2.3 App preview video (once screenshots are solid)

• detect: HYBRID — EV.meta / MANUAL
• rule: ✅ preview video live · 🟠 planned/produced but not shipped · 🔴 none and screenshots already solid · applies-if: screenshots done (else ⚪ — sequence matters)
• new-app: defer
• fix: app-store/screenshot-planner → /apple:screenshots

P2.4 Trust signals: privacy labels, support URL, account deletion — core

• detect: HYBRID — EV.meta (support/privacy URLs live) + EV.code.accountdeletion + MANUAL Q6 (labels reviewed)
• rule: ✅ all three verified current · 🟠 URLs live but labels unreviewed, or account deletion N/A-unclear · 🔴 broken URL or missing deletion where accounts exist
• new-app: plan
• fix: legal/privacy-publish + generators/account-deletion → /apple:privacy

P2.5 Alternate icons pre-shipped for future PPO icon tests

• detect: CODE EV.code.alticons
• rule: ✅ candidate alternate icons in the shipped asset catalog · 🔴 none (icon PPO tests blocked one full release)
• new-app: code — ship candidates in v1.0; unused alternates cost nothing
• fix: generators/product-page-optimization → /apple:icon

P3 — Localization

P3.1 Metadata-only localization for the big seven — core

• detect: MCP EV.locales
• rule: ✅ ≥7 of {ja, de, fr, es, pt-BR, ko, zh-Hans} localized (title/subtitle/keywords minimum) · 🟠 1–6 non-English locales · 🔴 English only
• new-app: plan — metadata-only is Level 0 localization; the app itself can wait
• fix: product/localization-strategy → /apple:localize

P3.2 Per-storefront ratings strategy

• detect: HYBRID — EV.territories (where it's sold) + MANUAL (prompting strategy per market)
• rule: ✅ aware ratings don't carry across storefronts AND prompting actively in each target market · 🟠 selling wide, prompting only at home · 🔴 assumed the home rating shows everywhere
• new-app: defer — activates with first non-home-market push
• fix: app-store/ratings-mechanics → /apple:localize

P3.3 PPP pricing for India, Brazil, Turkey, Indonesia

• detect: MCP EV.pricing + EV.pricepoints
• rule: ✅ manual price points set for IN/BR/TR/ID below auto-equalized levels · 🟠 some manual overrides · 🔴 pure auto-equalization · applies-if: paid app or IAP/subs sold in those territories
• new-app: plan
• fix: monetization (pricing-models, Pricing Localization) → ASC pricing (dry-run via /apple:ship)

P3.4 App itself localized where demand is proven

• detect: HYBRID — EV.sales/EV.analytics by territory vs EV.code.locales
• rule: ✅ in-app localization matches the markets producing revenue · 🟠 a market >15% of revenue still un-localized in-app · 🔴 localized nowhere despite proven foreign demand · ⚪ no foreign demand signal yet
• new-app: defer
• fix: product/localization-strategy → /apple:localize

P3.5 Metadata updates automated via the ASC API

• detect: CODE EV.code.automation
• rule: ✅ scripted/bulk metadata pushes (fastlane deliver, ASC API scripts, bulk tooling) · 🟠 partially scripted · 🔴 35+ locales maintained by hand
• new-app: plan
• fix: /apple:localize + /apple:metadata (bulk update paths) — standalone: _shared/asc-api

P4 — Ratings machinery

P4.1 requestReview at success moments — core

• detect: CODE EV.code.reviewprompt
• rule: ✅ requestReview called at a success moment with condition gating · 🟠 called but at launch/random (guideline-safe but conversion-poor) · 🔴 never prompts
• new-app: code
• fix: generators/review-prompt → /apple:build

P4.2 Negative reviews answered — core

• detect: MCP EV.reviews
• rule: ✅ 1–3★ reviews from the last 90 days have developer responses · 🟠 some answered · 🔴 none answered (updated ratings replace old scores — replies are recoverable stars)
• new-app: defer — activates with first reviews
• fix: app-store/review-response-writer → /apple:ratings (gated replies via the ASC API)

P4.3 Phased release + manual version release habit

• detect: MANUAL Q7
• rule: ✅ both habits (7-day phased rollout, manual release after approval) · 🟠 one · 🔴 auto-release, full blast (one bad build can nuke the rating before you wake up)
• new-app: plan — set on the first release
• fix: app-store/ratings-mechanics → /apple:ship (phased release step)

P4.4 Never reset the ratings summary — guardrail

• detect: MANUAL (only if a reset is being considered; default ✅)
• rule: ✅ no reset planned/performed · 🔴 reset planned or recently performed
• new-app: plan — note the rule before v2.0 temptation arrives
• fix: app-store/ratings-mechanics (read it before touching the toggle)