n8n Workflow: Google Maps Leads → Verified Emails → Google Sheets

This guide builds an n8n workflow that turns “dentists in Irvine” into a Google Sheet of leads with verified emails. It uses only standard n8n nodes: HTTP Request, Wait, If, Split Out, Filter and Google Sheets. You can run it on n8n Cloud or self-hosted.

The flow:

1. Start a LeadMerlin search job
2. Wait, then poll the job until it’s completed
3. Fetch all results, following pagination
4. Split the businesses into items and keep the ones with a verified, non-role email
5. Append or update rows in Google Sheets

Before you start

• A LeadMerlin API key (starts with lm_; the free plan includes 100 lead credits)
• An n8n instance
• A Google Sheet with a header row. We’ll use: name, owner, email, email_status, phone, phone_type, website, city, rating, reviews, cms, gbp_url

Create the credential once

In n8n, create a Header Auth credential:

• Name: x-api-key
• Value: your lm_… key

Every HTTP Request node below uses this credential. The key is never pasted into a node body.

Node 1: Trigger

Use a Manual Trigger while building, then switch to a Schedule Trigger if you want a weekly refresh. Add a Set node after it holding your inputs:

{ "keyword": "dentist", "location_codes": [1013883], "strictness": "moderate" }

1013883 is Irvine, CA. To find codes for other cities, call GET https://api.leadmerlin.com/api/v2/locations?q=your city once (it’s free) and copy the location_code. One job accepts up to five codes.

Node 2: Start the search

HTTP Request node:

• Method: POST
• URL: https://api.leadmerlin.com/api/v2/searches
• Authentication: Generic → Header Auth → your credential
• Send Body: JSON

{
  "keyword": "{{ $json.keyword }}",
  "location_codes": {{ JSON.stringify($json.location_codes) }},
  "strictness": "{{ $json.strictness }}"
}

The API answers 202 with a job_id. From this point you are charged one credit for each business that has a website. Listings without a website and off-niche listings are free.

Nodes 3–5: The polling loop

Enrichment crawls each business’s website, verifies emails and looks up phone line types, so a job takes a while. Typically that’s a minute or two, longer for big cities. Build a simple loop:

1. Wait node: 20 seconds.
2. HTTP Request node: GET https://api.leadmerlin.com/api/v2/jobs/{{ $('Start search').item.json.job_id }} with the same credential.
3. If node: {{ $json.status }} is equal to completed.
   • true → continue to results
   • false → connect back to the Wait node

Add a second condition so a failed job doesn’t loop forever: route status = failed to a Stop and Error node.

The status response is worth logging, because it explains the result count:

{
  "status": "completed",
  "total_unique": 2,
  "enriched_count": 2,
  "skipped_count": 0,
  "off_niche_count": 0,
  "category_facets": { "Plumber": 2 },
  "progress": 100
}

(That sample comes from the free demo endpoint at /api/v2/demo/jobs/x. Point the nodes at /api/v2/demo/... while building and nothing is charged.)

Node 6: Fetch all results with pagination

HTTP Request node:

• Method: GET
• URL: https://api.leadmerlin.com/api/v2/jobs/{{ $('Start search').item.json.job_id }}/results
• Query parameter: limit = 200

Open Options → Pagination:

• Pagination mode: Update a Parameter in Each Request
• Parameter: Query, name cursor, value {{ $response.body.next_cursor }}
• Pagination complete when: Other → {{ $response.body.has_more === false }}

Each page has the shape { businesses: [...], next_cursor, has_more }.

Node 7: Split Out

Split Out node on the field businesses. Each business becomes its own item.

Node 8: Keep only the leads you can email

Filter node with two conditions (AND):

• {{ $json.contact.all_emails.find(e => e.email === $json.email)?.status }} is equal to verified
• {{ $json.contact.all_emails.find(e => e.email === $json.email)?.is_role_based }} is false

email is the best address LeadMerlin picked for the business. contact.all_emails[] lists every address found, each with:

• status: verified, catch-all, invalid or unverified
• is_role_based: info@, office@ and similar
• is_free_provider: Gmail, Yahoo and so on

Loosen the filter if you also want catch-all addresses, which accept everything and can’t be confirmed mailbox by mailbox.

Node 9: Shape the row

A Set node maps the nested JSON to flat columns:

Node 10: Google Sheets

Google Sheets node → Append or Update Row:

• Column to match on: website (or gbp_url for businesses without one)
• Mapping: Map Automatically, since the Set node already uses your header names

Matching on website means re-running the workflow next month updates existing rows instead of duplicating them.

Variations worth building

• Qualify by tech stack. Add a Filter on web.tech.chat being empty or web.tech.cms being wix, which is useful if you sell chat widgets or websites. See how to qualify leads with tech stack signals.
• Enrich a list you already have. Swap Node 2 for POST /v2/enrichments with { "leads": [{ "domain": "example.com" }] } (up to 100 per request). It uses the same job polling and results flow.
• Send to a CRM. Replace the Sheets node with HubSpot, Pipedrive or an HTTP call to your outreach tool.
• Multi-city runs. Pass up to five location codes in one job instead of looping cities.

Costs and limits to keep in mind

• One lead credit per business with a website. Website-less and off-niche listings are free, and failed enrichments are refunded.
• If a search matches more businesses than your balance covers, it enriches the top-ranked ones your balance pays for and reports the rest as over_budget_count. Pass max_leads to cap a search yourself.
• Polling, locations and suggest calls don’t use credits, but a 20-second wait between polls is plenty.

Before you email anyone on the sheet, read our CAN-SPAM checklist for cold email to local businesses.