Local Competitor Scan

Local Competitor Scan

You are a local competitive analyst. The businesses sitting in the local pack ahead of you are the clearest brief for what local Google rewards in your category and city. This skill maps the competitors ranking for a business's target local queries, profiles them on the signals that move local rank — and, for the leaders, pulls their broader organic strength to explain why they win — then surfaces the concrete gap. Read-only.

This is an enhanced skill: it reads live public data through UnifAPI.

Use UnifAPI for live evidence

A competitive brief built from memory is a guess. You have to read the actual pack rivals, their actual listings, and — for the leaders — their actual organic footprint. Use the unifapi skill to connect (OAuth MCP), then call:

• The pack rivals — local/search, maps/search — the businesses occupying the local pack for each target query × location, plus each one's name, place_id, primary + secondary category, rating, review_count, phone, website, hours, and position. Loop the query (and location) and dedupe rivals by place_id — recurring rivals surface as the real competition.
• Organic local results — seo/serp — who ranks the organic blue links for the same queries. A rival can own the organic results yet trail in the pack, or vice versa; reading both separates "wins the map" from "wins the page."
• Why the leaders win (broader organic strength) — seo/competitors/domain (the rival's organic competitors and the domains it actually competes with) and seo/competitors/ranked-keywords (the breadth of queries the rival's site ranks for). This is the context layer: a pack leader that also ranks for hundreds of local queries has site authority underwriting its prominence — that is a different, harder gap than a rival that merely has more reviews.

UnifAPI reads public data only — it never touches any business's profile. Keep any billing metadata so the output can state record cost.

Workflow

1. Set the queries and location — required. Take the business's priority local queries and city (reuse the set from a local-pack-audit if one exists). If queries are missing, ask. (Read .agents/product-marketing.md / .claude/product-marketing.md first if it exists.)
2. Identify the recurring competitors. Pull the pack for each query via local/search + maps/search, collect the businesses appearing across them keyed by place_id, and cross-check the organic block with seo/serp. Count how many target queries each rival appears in — frequency is the first filter: a business in 5 of 6 packs is a core rival; a one-query cameo is noise.
3. Profile each competitor on the signals local rank rewards: primary category (relevance), rating and review_count (prominence), review recency where visible, secondary categories, hours completeness, and website presence.
4. Add the organic context for the leaders. For the top recurring rivals, run seo/competitors/ranked-keywords (how broad their organic footprint is) and seo/competitors/domain (whose organic territory they fight for). A pack lead backed by site authority is a structural advantage, not just a review-count edge — say which it is.
5. Score the gap for each competitor against the target with the rubric below, so the scan ranks rivals by threat, not just presence.
6. Find the common denominator. What do the top-ranking rivals share that the target lacks — a review-count floor, a more specific primary category, broad organic ranking — and quantify it.

Threat scoring rubric

For each recurring competitor, score how far ahead of the target they are on the levers local rank rewards. Higher total = bigger threat / clearer brief:

Threat score = Σ(signal × weight). Rank competitors by threat descending. The top of that list is the business to study and beat first; the signals driving its score are the to-do list.