Recipes / ICP (ideal customer) list to Salesforce: build a fit-checked prospect list in LeadIQ and export only new people

ICP (ideal customer) list to Salesforce: build a fit-checked prospect list in LeadIQ and export only new people

This recipe uses the official LeadIQ and Salesforce MCP servers, and both work in Claude and ChatGPT. Steps that write, send or spend credits wait for your approval.

IntermediateSDRs RevOps

Who is this for?

SDRs and RevOps people who build prospect lists from an ideal customer profile (ICP), meaning the kind of company and buyer you sell to best. You get the list into Salesforce with no duplicates and no clashes with accounts another rep owns. You can swap LeadIQ for Apollo, ZoomInfo or Cognism.

What do you get?

25 matches checked against Salesforce, 9 ready to export. Sample output: every company and person in it is fictional.

Credit check: 412 Universal Credits left. ICP used: VP or Director of Operations, logistics and freight, 250–2,000 employees, US West. LeadIQ matches: 25 profiles, about 2.5 UC (0.1 UC each). Saved to the LeadIQ list 'Ops West Q4' (free).

Person	Company	Fit	Salesforce
Dana Okafor	Larkspur Freightworks (fictional)	Strong. Title and size match	Contact exists, owner Priya Raman. Skipped
Sam Whitlock	Tidewell Robotics (fictional)	Strong. Hiring ops roles	No account. Ready
Alex Rivera	Mossgate Labs (fictional)	Weak. 90 employees	Below size. Dropped

Summary: 9 ready to export, 11 already in Salesforce (owned), 5 dropped on fit. Work emails for the 9: about 9 UC (1 UC each), plus about 0.9 UC to verify them (0.1 UC each). About 10 UC in all. Not looked up yet. Waiting for approval: look up and verify work emails for 9 people, then export them to Salesforce as Leads. Nothing has been exported yet.

How long does it take, and what does it cost?

Estimate About 20–30 minutes per list, most of it spent reviewing matches.

Assumptions: One list of about 25 matches, with work emails looked up and verified for about 10 people.

LeadIQ: A free credit check, one people search using your ICP (profiles cost 0.1 UC each, so about 2.5 UC for 25), and free saves to a prospect list. After you approve: one work email lookup (1 UC), one email verification (0.1 UC) and one Salesforce export per person. Some tools use credits. Uses Universal Credits (UC) per data point returned: profile 0.1 UC, work email 1 UC, direct phone 10 UC, company firmographics 3 UC, email verification 0.1 UC. List management, saved-prospect access and CheckCredits are free; re-lookups within a year are free. Exact rates vary by plan. Source: developer.leadiq.com, checked 2026-10-16
Salesforce: One search per person or company to find existing Leads, Contacts and Account owners. Nothing is written to Salesforce through this server. Every call uses quota. No paid credits; every tool call counts against the org's daily Salesforce API request quota (one or more API calls per tool). Source: github.com, checked 2026-10-16

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

Which MCP servers does it use?

Every server is confirmed in: ClaudeChatGPT

LeadIQ Up · auth

ClaudeChatGPTRead + writeSome tools use credits
https://mcp.leadiq.com/mcp
Salesforce Up · auth

ClaudeChatGPTCursorRead + writeEvery call uses quotaNeeds admin or IT
https://api.salesforce.com/platform/mcp/v1/platform/sobject-all

How do I set it 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.

Connect LeadIQ without code (Claude, ChatGPT)
Connect Salesforce without code (Claude, ChatGPT)

Developer setup

Auth setup and mcp.json for Claude Code and Cursor

LeadIQ: Sign in with OAuth in your client (run /mcp in Claude Code). Do it once interactively, and the client stores and refreshes the token.
Salesforce: 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": {
    "leadiq": {
      "url": "https://mcp.leadiq.com/mcp"
    },
    "salesforce": {
      "url": "https://api.salesforce.com/platform/mcp/v1/platform/sobject-all"
    }
  }
}

Step-by-step prompts

Check my LeadIQ credits. Each profile the search returns costs about 0.1 UC, so tell me the cost before you run it. Then find people who match this ideal customer profile (ICP): {titles}, {industry}, {employee_range} employees, in {location}. Show the first 25 with name, title, company, company size and a fit label (Strong or Weak) with a one-line reason. Don't look up emails or phone numbers, and don't save anything yet.
Save the people I keep to a LeadIQ prospect list called {list_name}. Then search Salesforce for each person (by name plus company) and each company (by website domain). Show who already exists as a Lead or Contact, and who owns any matching Account. Don't change anything in Salesforce.
Give me the final list: people with no Salesforce match and no owned Account. Show the credit cost of looking up and verifying their work emails. Only after I approve, look up the emails and verify them.
First, show me exactly who will be exported. Only after I approve, export those people from LeadIQ to Salesforce as {Leads_or_Contacts}, and report anything LeadIQ flags as a duplicate.

Can I run it on a schedule? (optional)

Run it by hand when you start a new campaign or territory. Building a list is a one-off job, not a daily one.
If you refresh the same list monthly, have the scheduled run do only the search and the Salesforce check, then stop. You approve email lookups and exports yourself.
Start with a small list (10 people) so you see the real credit cost before you run a big one.

Where does this go wrong?

Duplicates in Salesforce. The person may already be a Contact under a different email, or the company name may be spelled differently. Search Salesforce by name plus company and by website domain. Skip anyone who matches.
Exporting into someone else's account. An existing Account usually has an owner. List the owner and skip those people, or send the list to that owner. Don't create Leads on their account.
A loose ICP that wastes credits. Broad filters return lots of weak matches. Read the first 10 results, then tighten the title, company size or location before you save more or look up any emails.
People added as the wrong record type (Leads vs Contacts). Teams differ on whether new people go in as Leads or as Contacts on an Account. Confirm your team's rule before the export, and name your choice when you approve it.

Gotchas

LeadIQ charges Universal Credits (UC) for each piece of data it returns: 0.1 UC for a profile, 1 UC for a work email, 0.1 UC to verify an email and 10 UC for a direct phone number. Credit checks, list management and access to saved prospects are free. Exact rates vary by plan.
LeadIQ does the export to Salesforce, not the Salesforce server. This recipe uses Salesforce only to check what already exists.
An admin has to set up an External Client App before anyone can use Salesforce's hosted server, so a rep can't connect it alone. It also needs Enterprise Edition or above, and every call counts against your org's daily API limit.
Using HubSpot instead? LeadIQ's listing only names a Salesforce export. So after you approve, you'd check and create the records through the HubSpot connector, as in the champion job-change recipe.
In ChatGPT, LeadIQ needs a paid plan and Developer Mode. Salesforce needs Developer Mode too.

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.