THE CLAUDE CODE LEAD ENRICHER

How to enrich your account list with verified contact data without paying $500/month.

This system replaces your Clay subscription with a direct connection between Claude Code and your enrichment provider.

You'll set it up in under 30 minutes, and from that point on, finding the right people inside your target accounts and pulling their verified contact data costs a fraction of what Clay charges.

The Founder’s GTM 🦊 paid subscription includes this system plus others. You can cancel anytime and keep everything you've already accessed.

Table of Contents


Introduction

  • What this guide does

  • Who this is for

  • What you will have by the end

  • What you need before you start


Part 1: How MCP Enrichment Works

  • Why Clay became the default enrichment tool

  • What MCP changes about the enrichment equation

  • The direct connection model: five steps from setup to output

  • What this stack costs vs what Clay costs

  • An important distinction: provider-verified vs currently deliverable

  • What this stack does not replace

  • The MCP ecosystem is growing


Part 2: Choosing Your Enrichment Provider

  • What to look for: five criteria that actually matter

  • Coverage for your specific ICP

  • Email verification quality

  • MCP server availability and quality

  • API pricing at founding team scale

  • Rate limits and session behaviour

  • The three main options:

    • Apollo: firmographic matching at scale, official MCP server

    • Prospeo: European ICP coverage, real-time verification

    • Findymail: LinkedIn-first, strong for European B2B professionals

  • The coverage gap: when you need more than one provider

  • Running the coverage test

  • What the numbers mean: 80%, 60% to 80%, below 60%

  • The most common fallback combinations

  • Making your decision


Part 3: Setting Up Your Enrichment Provider

  • Before you start: create your project folder

  • Setting up Apollo

    • Step 1: Get your Apollo API key

    • Step 2: Connect Apollo via MCP

    • Step 3: Verify the Apollo connection

  • Setting up Prospeo

    • Step 1: Get your Prospeo API key

    • Step 2: Connect Prospeo via MCP

    • Step 3: Verify the Prospeo connection

  • Setting up Findymail

    • Step 1: Get your Findymail API key

    • Step 2: Connect Findymail via MCP

    • Step 3: Verify the Findymail connection

  • Verifying multiple providers are connected

  • Running your coverage test

  • Adding a second provider as a fallback

  • Troubleshooting: provider does not appear in the MCP tools list

  • Troubleshooting: authentication error when provider is queried

  • Troubleshooting: provider connects but returns empty results


Part 4: Preparing Your Account List

  • The required column structure

  • Pre-enrichment columns: company_name, domain, industry, employee_count, geography, buying_signal, why_fit_evidence

  • Enrichment columns: champion and buyer contact fields, news_summary, tech_stack_signals, research_notes, source_url, date_enriched

  • Status columns: research_done, status

  • Adding the enrichment columns to your CSV

  • The complete header row

  • Cleaning your domain column

  • Validating ICP fit before enrichment

  • Prioritising your list with priority_targets.txt

  • The five-minute pre-enrichment check

  • A note on list size and batch quality


Part 5: The /enrich Skill

  • What the Skill does and why it matters

  • Creating the Skill file: full template

    • BEFORE YOU START: load CLAUDE.md, identify unprocessed rows

    • YOUR ROLE: quality over volume, no invented data

    • STEP 1: Clarify the session (three questions before searching)

    • STEP 2: State your enrichment plan

    • STEP 3: Find the Champion contact (primary query, fallback logic)

    • STEP 4: Find the Buyer contact (same person handling, fallback logic)

    • STEP 5: Verify and score the row (verification pass, scoring pass)

    • STEP 6: Save to companies.csv (column rules, save frequency)

    • STEP 7: End of session report (seven sections)

  • How to run the Skill

  • What a fully enriched row looks like

  • Adapting the Skill to your provider


Part 6: The Decision Guide

  • The question to ask first: what job is Clay actually doing?

  • Discovery 1: Clay is doing one job you are paying for three

  • Discovery 2: Clay is doing exactly what you need

  • Discovery 3: Clay was solving a problem that no longer exists

  • When the direct MCP stack is enough

    • Your ICP is well-covered by one or two providers

    • You are the primary person running enrichment

    • You value research quality over raw contact volume

    • You want full control over your data

  • When Clay is still worth keeping

    • Waterfall enrichment across five or more providers at scale

    • Your team is larger than three or four people

    • You rely on Clay’s native outreach tool integrations

    • Your enrichment logic is highly complex

  • The hybrid stack: Claude Code for research, Clay for enrichment

  • A practical cost comparison: direct stack vs hybrid stack vs Clay-only

  • Making your decision: three questions


Part 7: Troubleshooting

  • Category 1: Connection Problems

    • Problem 1.1: Provider does not appear in the MCP tools list

    • Problem 1.2: Authentication error when provider is queried

    • Problem 1.3: Provider connects but returns empty results

    • Problem 1.4: MCP server crashes mid-session

  • Category 2: Data Quality Problems

    • Problem 2.1: Verified email hit rate is lower than expected

    • Problem 2.2: Provider returns contacts for the wrong company

    • Problem 2.3: Contact information is outdated

    • Problem 2.4: news_summary and tech_stack_signals fields are consistently empty

  • Category 3: Session Behaviour Problems

    • Problem 3.1: Session times out before completing the batch

    • Problem 3.2: Skill processes already-enriched rows again

    • Problem 3.3: Scoring is inconsistent between sessions

  • Category 4: CSV Problems

    • Problem 4.1: Skill creates new columns instead of writing to existing ones

    • Problem 4.2: CSV becomes corrupted after a session

    • Problem 4.3: File grows too large for efficient processing

  • When to ask for help: the diagnostic prompt


Part 8: What Comes Next

  • What you have built: identity, contacts, context, prioritisation

  • How this list feeds into the outreach sequence

    • why_fit_evidence and buying_trigger feed the WHY THEM hook

    • tech_stack_signals feed the WHY YOU proof point selection

    • champion_email and buying_signal feed channel and sequence decisions

    • research_notes feed the manual review step

  • The next guide: building the outreach sequence with Claude Code

  • Three things to do before the next guide

  • The compounding effect

  • Final checklist before the next guide

Keep reading with a 7-day free trial