Competitor teardown

Competitor Teardown

Answer the question behind the question: not "what are X's numbers" but "where is X vulnerable and what should we copy vs. counter".

Requirements

GetAppNiche MCP connected: search_apps (to resolve names to ids), get_app_detail, get_app_historicals, get_app_reviews. Canonical ids look like apple:284882215 / google:com.example.app, but get_app_detail and get_app_reviews also accept a bare Apple track id, an Android package name, or a store URL pasted from a browser — so when the user hands you a link, use it directly instead of searching for the app.

Workflow

1. Resolve and confirm. If the user gave a link or an id, skip straight to step 2. Otherwise find each competitor with search_apps (search_in: "title" narrows a name lookup); if multiple plausible matches, confirm with the user before spending credits on the wrong app.
2. Snapshot each app with get_app_detail: downloads and revenue, rating and review count, price model, category, last update.
3. Pull trajectory with get_app_historicals (90 days is a good default; up to 730 available). Look for: growth vs. plateau vs. decline, revenue moving independently of downloads (monetization change), rating drift (quality trend).
4. Read positioning from metadata. From each app's title/subtitle/description: who they target, the promise they lead with, keywords they visibly chase.
5. Hear their users. Call get_app_reviews with the competitor's app_id — the live lookup works on any app in either store (Apple id or Android package name), no monitoring needed. One page of recent reviews turns "weaknesses worth exploiting" from inference into evidence: what their users actually complain about, and which complaints track the latest version.
6. Synthesize per competitor: two or three strengths worth copying and two or three weaknesses worth exploiting — each tied to a number, a quote, or an observed fact, not vibes.
7. If the user has their own app, finish with a positioning gap: the audience or promise none of the competitors own, and the shortest credible path to owning it.

Output

Per competitor: a compact stat block (downloads · revenue · rating · trajectory arrow) plus the strengths/weaknesses lists. Then a comparison table if 2+ competitors, and the gap recommendation.

Guardrails

• Round figures for readability — say "~$40K/mo", never a false-precision figure.
• Historicals cost 1 credit per app; for a 5+ app comparison, snapshot all first and pull history only for the 2–3 that matter.
• A teardown without a recommendation is a report — always end with "so what".