Local & Places

Local AI Visibility Audit

Reveal which local businesses get recommended for 'best X in city' across Google's AI and the community threads it cites — and which strong, well-reviewed ones stay invisible

What it does

Google's AI Overviews and AI Mode now decide who gets recommended for 'best X in city,' and they pull heavily from Reddit and forum threads. Optionally union those community sources with the AI-cited set, diff the whole recommendation landscape against the real high-rated businesses on Maps, and you surface strong companies that are invisible to AI — a clean sales wedge. Which community platforms to run, the country, and how deep to pull Maps results are all tunable.

Who it's for: Local SEO and AI-visibility agencies prospecting for work

Inputs

How your agent runs it

1. 1
   google_ai_mode

   google_ai_mode(prompt="best {category} in {city}")

   Capture reply_parts and reference_links to record exactly which businesses Google's AI Mode names and cites. country is optional — omit it to let Google infer the region.

2. 2
   search_web

   search_web(query="best {category} {city}", include_ai_overview=true, location="{city}")

   Pull ai_overview.text_parts and reference_links. location="{city}" simulates a searcher inside the metro so the AI Overview reflects true local results. Union with step 1 to build the AI-recommended set.

3. 3
   search_reddit

   search_reddit(query="best {category} {city}", sort_by=top)

   Only run if {platforms} includes reddit: surface the businesses the community recommends in 'best {category} {city}' threads — a major source AI Overviews cite. Add these as a second recommendation set.

4. 4
   search_forums

   search_forums(query="best {category} {city}", time=year)

   Only run if {platforms} includes forums: capture business names recommended in independent forum threads over the past year to widen the community-recommendation set beyond Reddit.

5. 5
   search_places

   search_places(query="{category} {city}", pages=10)

   List every OPERATIONAL business with strong rating and review_count — the ground-truth set that deserves to rank. pages={result_pages} (default 10) controls how deep to pull.

6. 6
   place_details

   place_details(place_id=<place_id>)

   For high-rated businesses absent from BOTH the AI set and any community set, pull contact info (phone, website, email if present) to pitch an AI-visibility fix.

Tools used

The full playbook

This is exactly what the MCP returns to your agent (via the local-ai-visibility-audit prompt or get_skill tool), with your inputs filled in.

SKILL: Local AI Visibility Audit
Google's AI Overviews and AI Mode now decide who gets recommended for 'best X in city,' and they pull heavily from Reddit and forum threads. Optionally union those community sources with the AI-cited set, diff the whole recommendation landscape against the real high-rated businesses on Maps, and you surface strong companies that are invisible to AI — a clean sales wedge. Which community platforms to run, the country, and how deep to pull Maps results are all tunable.

You are running this skill on API Direct via its MCP tools. Execute the steps below yourself by calling the named tools in order — values in <angle brackets> come from a previous step. Then deliver the result described at the end.

INPUTS:
  - category (required): <category — ASK THE USER>
      Business category to audit
  - city (required): <city — ASK THE USER>
      City and region to audit
  - platforms (optional): (optional — e.g. reddit, forums)
      Comma-separated list of platforms to run — options: web, reddit, forums, places. Omit to run all of them; name specific platforms to limit the run.
  - country (optional): (optional — e.g. us)
      Two-letter country code to localize all Google results (AI Mode, AI Overview, Maps, place details). Omit to let Google infer region.
  - result_pages (optional): 10 (default — override if you like)
      How many pages of Maps results to pull for the ground-truth set (default 10).

PLATFORM SELECTION: some steps are gated with "Only run if {platforms} includes X". If the user supplied a `platforms` value, run only the steps whose platform is listed. If the user did NOT supply `platforms`, run ALL steps — every platform the skill supports.

STEPS:
  1. Tool `google_ai_mode` — google_ai_mode(prompt="best {category} in {city}")
     Capture reply_parts and reference_links to record exactly which businesses Google's AI Mode names and cites. country is optional — omit it to let Google infer the region.
  2. Tool `search_web` — search_web(query="best {category} {city}", include_ai_overview=true, location="{city}")
     Pull ai_overview.text_parts and reference_links. location="{city}" simulates a searcher inside the metro so the AI Overview reflects true local results. Union with step 1 to build the AI-recommended set.
  3. Tool `search_reddit` — search_reddit(query="best {category} {city}", sort_by=top)
     Only run if {platforms} includes reddit: surface the businesses the community recommends in 'best {category} {city}' threads — a major source AI Overviews cite. Add these as a second recommendation set.
  4. Tool `search_forums` — search_forums(query="best {category} {city}", time=year)
     Only run if {platforms} includes forums: capture business names recommended in independent forum threads over the past year to widen the community-recommendation set beyond Reddit.
  5. Tool `search_places` — search_places(query="{category} {city}", pages=10)
     List every OPERATIONAL business with strong rating and review_count — the ground-truth set that deserves to rank. pages=10 (default 10) controls how deep to pull.
  6. Tool `place_details` — place_details(place_id=<place_id>)
     For high-rated businesses absent from BOTH the AI set and any community set, pull contact info (phone, website, email if present) to pitch an AI-visibility fix.

DELIVER: A gap report of strong local {category} businesses absent from Google's AI (and, if selected, the Reddit/forum threads it cites), each with contacts for an AI-visibility pitch

Note: each underlying tool call is billed at its normal endpoint price; get_sentiment adds a small per-page surcharge. Page through results as needed but stop once you have enough to deliver the outcome.