- Home
- Skills
- Sales & Lead Gen
- Local Lead Prospector — Google Maps B2B Lead Scoring System
Works with the AI Agents you already use
Local Lead Prospector — Google Maps B2B Lead Scoring System
Turn Google Maps into a prioritized B2B call list: your agent finds local businesses by trade and area, scores fit, size and phone contactability, dedupes them into one master list.
$9.99
Local Lead Prospector — Google Maps B2B Lead Scoring System
Example session with this skill installed
I sell maintenance contracts for commercial kitchens. Find me restaurants and bakeries in Downtown and Riverside — about 50 callable leads. Keep the master list in leads/master-leads.csv. (Fictional example.)
- Read your context and instructions
- Compiled the local lead prospector
Illustrative run with fictional businesses:
Plan: 4 searches (restaurants x Downtown/Riverside, bakeries x Downtown/Riverside), cap 60 per search.
Run complete: 212 businesses found - 148 new - 61 duplicates skipped - 3 blocklisted skipped.
Master list: 148 leads — 39 HIGH - 57 MEDIUM - 121 callable (phone listed).
Call list — top of HIGH:
- Hill Country Diner (fictional) — Restaurant, Downtown — 83 pts — Target fit +20 - 240 reviews +30 - Phone +25 - Rating 4.7 +8
- Riverside Bakehouse (fictional) — Bakery, Riverside — 79 pts — Target fit +20 - 132 reviews +30 - Phone +25 - Rating 4.2 +4
- Casa Verde Kitchen (fictional) — Restaurant, Downtown — 63 pts — Target fit +20 - 58 reviews +18 - Phone +25 - Rating 3.9 +0
Gaps: Bakeries in Riverside came back thin (2 callable targets). Suggested: add "cafes in Riverside", or extend coverage to Northgate.
Held back: 27 leads have no listed phone — they are in the master list, out of the call list.
Connects securely to your tools. The creator never sees your data.
About this skill
This is a local B2B prospecting system — not a scraping tool and not a contact database. Nothing is pre-collected and no code ships with it: the system directs your own AI agent to work Google Maps in your own browser session, business by business, at a human pace — and turns what it finds into a call list where every lead explains its own score.
What it does
- Starts from your offer. You say what you sell; the system helps you define the business types that buy it (segments with keyword lists) and the areas to cover, then builds a query plan — one
"<trade> in <area>"search per segment × area. - Searches Google Maps like a researcher, not a bot. Two phases per query: collect candidates from the results list, then open one business at a time and record name, category, rating, review count, phone, website and address — copied exactly as shown. If Google asks for verification, the run stops. Missing phone or website is recorded as "not listed", never invented.
- Classifies target fit. Each business is matched against your segment keyword lists (stems, first match wins, checked against name + category + the query that found it).
- Scores every lead — and shows its work. Fit +20 · size by review count (≥100 → +30, ≥50 → +18, ≥20 → +8) · phone listed +25 · rating (≥4.5 → +8, ≥4.0 → +4). Max 83. HIGH ≥ 55, MEDIUM ≥ 30. Every lead stores its score breakdown, so you always know why it's ranked where it is.
- Enforces the fit gate. A business outside your target segments is always LOW priority — even a huge, highly-rated one with 59 points. Your calling time goes to plausible buyers first.
- Never duplicates. Candidates are checked against your master list before being opened, and again before saving. Repeat runs only add new businesses; re-encounters just update a "last seen" date.
- Protects your work. One persistent master list (CSV or spreadsheet) with your own outreach columns — status, contacted, follow-up, notes — that the system creates empty and never overwrites, in any mode. A blocklist keeps rejected businesses out forever.
- Refresh mode. Changed your segments or scoring weights? The system re-classifies and re-scores your entire existing list from stored data — no new searches — and reports how many leads changed tier.
- Delivers a call list plus the truth about coverage. HIGH first, callable leads only, each with phone, score and reason — and a summary that names the gaps: segments or areas where coverage is still thin, with suggested next searches.
The scoring model
| Signal | Condition | Points |
|---|---|---|
| Target fit | Matches one of your segments | +20 |
| Size proxy | 100+ reviews | +30 |
| Size proxy | 50–99 reviews | +18 |
| Size proxy | 20–49 reviews | +8 |
| Contactability | Phone number listed | +25 |
| Quality signal | Rating 4.5 or higher | +8 |
| Quality signal | Rating 4.0–4.4 | +4 |
Review count is used as a size proxy: for local businesses, more reviews over time means a bigger, more established operation. A listed website is recorded but deliberately not scored — it doesn't predict B2B fit; a callable phone does. All weights and thresholds are configurable, and refresh mode re-scores your whole list when you change them.
Who it's for
- B2B sellers whose buyers are local businesses: suppliers, distributors, service contractors, agencies, wholesalers, equipment and maintenance providers.
- Sales reps opening a new territory who need to know who to call first, not just who exists.
- Small teams that prospect repeatedly and are tired of rebuilding the same spreadsheet — and of duplicates.
What's included
SKILL.md— the complete system: inputs → query plan → Maps search protocol → classification → scoring → deduplication and master list → refresh mode → call-list delivery.- Setup guide — requirements, how to design segments and keyword lists that classify well, master-list anatomy, scoring tuning for your market, troubleshooting.
- Startup-behavior reference — drop it into your project so your agent opens every session with the 5 setup questions and this workflow ready.
Requirements
An AI agent that supports the SKILL.md standard (Claude Code, Codex, Cursor and others) with browser access, and a folder or spreadsheet for the master list. No API keys, no database, no code to install.
Proven in production
This system generalizes a prospecting pipeline that ran in production for a B2B supplier in Mexico, where it built and maintained a scored, deduplicated master list of local business leads used for real outbound sales. This edition is fully generalized: you define your own offer, segments, areas and scoring — no data, names or defaults from that deployment are included.
Honest notes
- Instructions only. The package contains no software and no pre-made list. Your agent does the work by browsing, in your session, at a human pace — expect a research session, not an instant export of thousands of rows.
- Public data only, recorded faithfully. If a business doesn't publish a phone, the lead says "not listed". The system never guesses contact details.
- It builds lists; it never contacts anyone. Outreach — and complying with your jurisdiction's anti-spam, telemarketing and privacy rules — is yours.
Instructions only: the package contains no software and no pre-made list — expect a research session at a human pace, not an instant bulk export. Public data only: a business without a listed phone is recorded as "not listed", never guessed. The system builds lists; it never contacts anyone — outreach and local anti-spam/telemarketing compliance are yours.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 day ago
Requires
Permissions
No special permissions declared or detected