La interfaz está localizada; las descripciones y documentos de las Skills se traducen automáticamente con IA a su idioma. El código, los comandos, rutas y términos técnicos se mantienen en el original, y prevalece el documento del autor.
Explorar Skills
Datos e investigación · SKILL.md

plan-aso

plan-aso

Plan-only ASO audit of App Store and Google Play listings: keywords, localized metadata, screenshots, ratings prompts. Use when "optimize our store listing", "improve app downloads", or "ASO". Submission readiness → plan-mobile-readiness.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentación de la Skill

GitHub

name: plan-aso description: > Plan-only ASO audit of App Store and Google Play listings: keywords, localized metadata, screenshots, ratings prompts. Use when "optimize our store listing", "improve app downloads", or "ASO". Submission readiness → plan-mobile-readiness. license: MIT effort: high

plan-aso — Store listing growth plan

Degree of freedom: HIGH — audit the listing, emit a plan. Stay plan-only. No metadata, screenshot, or prompt edits until approved.

Role: Mobile growth / ASO specialist.

Task: Audit the listing that decides whether a visitor installs, emit plan-aso.md. Plan only — no metadata, screenshot, or prompt edits until approved.

ASO is two funnels: keywords decide who finds the app; screenshots + first impression decide who installs.

This skill vs neighbors

SkillOwns
plan-aso (this)Find + install conversion of the listing
plan-mobile-readinessSubmission mechanics, privacy manifest, demo account
plan-privacy-complianceHonest privacy labels (don't ASO-lie about data)
enhance-web-seo / plan-aeo-readinessWeb / answer-engine discoverability
audit-monetization-iapWhether IAP actually works
design-frontendAsset production after the plan is approved

Do not fire for "will Play reject us / Guideline 2.5.2" → plan-mobile-readiness.

How to reason (every plan item)

  1. Propose — keyword, locale, screenshot order, or ratings-prompt change
  2. Risk — find vs install: who never sees the app, or who sees it and bounces
  3. Keep-working — locales/assets that already convert
  4. Phase — quick / medium / ongoing (do not execute)

Worked example

Propose: fill the unused 40 chars of the iOS keyword field; drop title-word repeats; put the benefit screenshot first. Risk: core use-case terms never indexed; first impression is a bare UI dump. Keep-working: JP long description already locale-native. Phase: quick — keywords + screenshot reorder. Find vs install: find (keywords) + install (shot 1).


Phase 0 — Gather the current listing [HIGH freedom]

Per platform × locale: name, subtitle (iOS) / short description (Android), iOS 100-char keyword field, long description, screenshots + captions, preview video, icon, category, ratings state. Note machine-translated or EN-only locales.


Phase 1 — Discoverability (find) [HIGH freedom]

iOS keywords — 100 chars fully used, comma-separated, no wasted spaces, no repeats of title/subtitle words, no plural+singular waste. Core use-case terms present.

Title & subtitle — Title carries the #1 keyword beyond brand. Subtitle is a keyword-bearing value line, not a slogan.

Android — Play indexes the long description. Core terms in the first lines, natural density, not stuffed.

Localization — Each locale is a separate keyword surface. JP + EN need locale-native keywords, not translated English. Untapped locales are the cheap win.

Category & competitors — Right primary category? Terms competitors rank that this listing ignores?


Phase 2 — Conversion (install) [HIGH freedom]

Screenshots — First 2–3 (no scroll) lead with benefit captions, not bare UI. Narrative across the set. Localized text for key markets.

Icon — Distinct at thumbnail size.

Preview video — Value in the first seconds (often muted autoplay).

Above-the-fold copy — One-line hook, not "Welcome to…".

Ratings — In-app prompt at a positive moment (after value, not on launch). Reviews responded to.

Trust — No leftover screenshots from an old version.


Phase 3 — Prioritized plan (approve before executing) [HIGH freedom — plan only]

  • Quick — fill keyword field, drop redundant title words, add missing core-market locale, reorder screenshots value-first
  • Medium — benefit captions, localized screenshot text, preview video, well-timed rating prompt
  • Ongoing — locale expansion, review cadence, store A/B on icon/shots, keyword iteration

Each item: current → change → find vs install → effort. Asset work flagged for design-frontend.

Self-critique before the burndown [LOW freedom — do not skip]

  1. Per platform × locale — do not audit EN-only and call JP done
  2. Find vs install labeled — every item says which funnel it moves
  3. No ASO-lie — privacy labels stay honest (plan-privacy-compliance)
  4. Right owner — Guideline 2.5.2 / demo account → plan-mobile-readiness
  5. Nothing edited on the listing until approved

Definition of Done

  • Listing inventory per platform × locale
  • iOS keywords + title/subtitle audited
  • Android description placement checked
  • Localization gaps with locale-native keyword notes
  • Screenshots / icon / video / hook / ratings reviewed
  • Plan prioritized; approved before listing edits

Output format

  1. Inventory — field × platform × locale
  2. Find findings — issue | severity | fix
  3. Install findings — issue | severity | fix
  4. Plan — quick / medium / ongoing
  5. Recap and await approval — end the turn with a standalone recap in chat: the two or three highest-impact findings and the first phase to approve. plan-aso.md is the deliverable; write it before the recap.

Related

  • plan-mobile-readiness — get into the store
  • plan-privacy-compliance — labels must match collection
  • audit-monetization-iap — paid features actually unlock
  • enhance-web-seo / plan-aeo-readiness — web counterparts
  • design-frontend — produce approved assets
Skills guardadas