Recipes / Find lookalike accounts and enrich them

Find lookalike accounts and enrich them

Apollo + Clay + HubSpot

IntermediateRevOpsSDRsMarketing

Who it's for

GTM engineers and SDR leads who build account lists from closed-won customers and want enriched, de-duplicated records in the CRM.

What you get

Lookalike accounts, ready for review (table before anything is saved). Sample output: every company and person in it is fictional.

Profile of your last 10 closed-won: B2B software, 80–300 employees, US or Canada,
usually a RevOps lead involved in the deal.

| # | Company                | Employees | Country | Fit | Contact (title)                    |
|---|------------------------|-----------|---------|-----|------------------------------------|
| 1 | Fernhill Software      | 140       | US      | 92  | Dana Okafor (Head of RevOps)       |
| 2 | Quarry Lane Analytics  | 210       | CA      | 88  | Sam Whitlock (Sales Ops Manager)   |
| 3 | Larkspur Freightworks  | 95        | US      | 81  | none found, skip or retry later    |
...17 more

Skipped 6 domains already in HubSpot. Nothing has been created yet.
Reply 'approve 1-2, 4-20' to create these companies and contacts.

Time and cost

Estimate About 15–25 minutes, including your review of the table.

Assumptions: One run with a 10-deal profile, 50 lookalike companies searched, and 20 people enriched.

Exact credit costs depend on your plan with each vendor. Check the tool pages before a large run.

MCP servers

Every server is confirmed in: ClaudeChatGPT

Set up in a chat app

Every server in this recipe is confirmed for Claude and ChatGPT. Connect each one, then paste the prompts below into one conversation.

Developer setup

Auth setup and mcp.json for Claude Code and Cursor
  • Apollo: Use OAuth sign-in for interactive use. For headless runs, use an API key in the X-Api-Key header.
  • Clay: Sign in with OAuth in your client (run /mcp in Claude Code). Do it once interactively, and the client stores and refreshes the token.
  • HubSpot: Sign in with OAuth in your client (run /mcp in Claude Code). Do it once interactively, and the client stores and refreshes the token.

Put this in .mcp.json (Claude Code, project root) or ~/.cursor/mcp.json (Cursor). Swap the example keys for your own, and don't commit real ones. Remove the // comment lines if your client requires strict JSON.

{
  "mcpServers": {
    "apollo": {
      "url": "https://mcp.apollo.io/mcp"
      // headless alternative: "headers": { "X-Api-Key": "YOUR_APOLLO_API_KEY" }
    },
    "clay": {
      "url": "https://api.clay.com/v3/mcp"
    },
    "hubspot": {
      "url": "https://mcp.hubspot.com"
    }
  }
}

Step-by-step prompts

  1. Pull the 10 most recent closed-won companies from HubSpot with industry, employee count and country. Summarize what they have in common.
  2. Using that profile, search Apollo for 50 companies that match. Exclude any domain that already exists as a company in HubSpot.
  3. For the top 20 by fit, run Clay enrichment to find the head of RevOps or sales ops at each. Show me the results as a table before saving anything.
  4. After I approve the table, create the companies and contacts in HubSpot and add a note saying the source is a lookalike search.

Run it on a schedule (optional)

Where this goes wrong

Gotchas

Download

Get the prompts, guardrails and expected output as one Markdown file. Paste it into a Claude Project or ChatGPT project's instructions, or save it as CLAUDE.md for Claude Code.

Download lookalike-accounts-enrich.md

Vendor sites: Apollo · Clay · HubSpot. If we join an affiliate program, these links will be labeled as affiliate links.