How to find local business decision makers that don't appear in B2B databases

Why this method works

Traditional B2B databases like Apollo and Sales Navigator miss 90%+ of local/service-based businesses because:

• Many don't have LinkedIn company pages
• Employees don't list these companies on their profiles
• Example: Apollo found only 1,300 UK clinics vs. 19,220 using this method

Results from this approach:

• 32% LinkedIn connection rate
• 24% reply rate
• 31 positive responses from ~1,000 outreach attempts
• 35 positive replies from 1,500 email campaigns

Tools required

• Apify Google Maps Scraper - to get data from Google Maps
• Clay - data enrichment and workflow
• n8n - automation workflow
• ChatGPT - generating city lists
• Lead Magic - personal email finder
• HeyReach - LinkedIn outreach
• Instantly - email outreach

Part 1: Getting Google Maps data

Step 1: Generate your location list

Get major cities in your target country:

1. Go to ChatGPT
2. Prompt: "List all major cities in [YOUR COUNTRY]"
3. Copy the list

Step 2: Set up Apify to get data from Google Maps

Basic configuration:

1. Choose category search (not search terms)
2. Select all relevant categories for your niche meticulously
   • Example: For clinics - select every clinic-related category
3. Set location to City + Country format (not just country)
   • Better results: "London, United Kingdom"
   • Worse results: "United Kingdom"

Why use categories? Captures businesses that don't use standard keywords in their names.

Part 2: Automate getting data in bulk (Optional)

Manual option: Takes ~1 hour to copy/paste locations individually

Automated option using Clay + n8n:

Step 3: Set up Clay table

1. Import your city list from ChatGPT into Clay
2. Create HTTP API column to trigger Apify runs
3. Get your Apify settings:
   • Configure one scraper run manually in Apify
   • Click "JSON" at the top
   • Copy the entire JSON object
   • Paste into Clay HTTP API column
4. Change only the location field to dynamically reference your city column
5. Run in batches (based on your Apify plan limits)

Step 4: Build n8n automation

Purpose: Consolidate all scraper runs into one file

1. Create webhook to receive data from Clay
2. Add "Get Run Data" node using the Dataset ID from Clay
3. Find API endpoint: Apify → API → API endpoints
4. Send consolidated data back to Clay

Why this workflow?

• Clay processes row-by-row (perfect for multiple locations)
• n8n handles the data consolidation
• Automatic data aggregation from all scraper runs

Part 3: Qualify and filter your list

Step 5: Filter out unqualified leads

Critical qualification prompts in Clay:

1. Exclude irrelevant business types

Example: NHS vs. Private Clinics

• Search their website
• Return: "NHS" or "Private"
• Filter out NHS (government-linked, lower budget)

2. Identify relevant software/tools they useCheck booking page URLs for:

• Cliniko (clinic management software)
• E-prescriber
• Other relevant tools

Why this matters? Software usage = positive buying signal for your offer.

Step 6: Find contact emails

Use Clay to scrape:

• Generic email addresses (info@, contact@)
• Useful for small clinics with single contact points
• Fallback option when personal emails aren't available

Part 4: Extract decision makers

Step 7: Scrape website for Staff Names

The money prompt - adapt this for your niche:

Visit this company website. Check the About Us page and Team page.Return a list of [DECISION MAKERS] with:

• Full name
• Job title
• Contact email (if available)Focus on: [YOUR TARGET ROLES - e.g., doctors, clinic managers]

Results example with this prompt: Found 41 doctors at one clinic vs. 2 on Apollo

This data lives on websites but not LinkedIn/Apollo - completely untapped.

Step 8: Transfer these results to new Clay table

Use "Write to Other Table" function:

1. Create new blank table
2. In original table: Add "Write to Other Table" column
3. Select your doctor/staff list (JSON array)
4. Check "Select from a list" - this creates one row per person
5. Map ALL data points you need:
   • Full name from JSON
   • Job title from JSON
   • Email from JSON
   • Company name (from parent table)
   • Company website (from parent table)
   • All other relevant fields

⚠️ Common mistake: Assuming all data transfers automatically - you must manually map each field.

Part 5: Cleanse and enrich contact data

Step 9: Extract and format names

Problem: Full names include titles like "Dr. Rita Johnson"

Solution: Clay prompt:

Extract first name from this full name: [FULL NAME]Rules:

• Remove titles (Dr., Mr., Mrs.)
• If only initial available (Dr. R. Johnson), return "Dr. Johnson"
• Return only the first name for personalization

Step 10: Categorize and re-qualify roles

Second qualification layer prompt:

Review this data:

• Job title: [JOB TITLE]
• Full name: [FULL NAME]
• Category: [CATEGORY]

Is this person a [TARGET ROLE]? Yes/No

Aim for 95% accuracy- better than manual work

⚠️ Critical: Use API keys (OpenAI, Anthropic) not Clay credits - much cheaper

Step 11: Enrich with contact details

Waterfall enrichment strategy:

1. Find LinkedIn URL first:Search prompt:Find LinkedIn profile for [FIRST NAME] [JOB TITLE] at [COMPANY NAME]Return only the URL, no other text
2. Personal email enrichment:
   • Tool 1: [First personal email tool]
   • Tool 2: [Backup personal email tool]
   • Requires LinkedIn URL to work effectively
3. Work email enrichment:
   • Use Lead Magic or similar
   • Expect low match rates for local businesses
   • That's why personal emails are critical
4. Fallback options:
   • Generic company emails from Step 6

Part 6: Launch outreach campaigns

Step 12: Export to outreach tools

You now have:

• 30,000+ qualified leads (vs. 1,300 from Apollo)
• Personal emails
• Work emails (where available)
• Generic company emails
• Personalization variables (software used, role, etc.)

Add to campaigns:

• LinkedIn: HeyReach
• Email: Instantly

Step 13: Write targeted copy

Personalization variables available:

• First name
• Company name
• Software they use (Cliniko, etc.)
• Role/specialty
• Location

Keep copy relatively generic when data is limited, but use specific callouts when you have them (e.g., "I noticed you use Cliniko...")

Expected results

Based on ~1,000-1,500 outreach attempts:

LinkedIn (HeyReach):

• 32% connection acceptance
• 24% reply rate
• 31 positive responses

Email (Instantly):

• 35 positive replies
• High personal email usage indicates data quality