AI Stack for Lead Enrichment

Clay + Clearbit + Apollo → Claude → Salesforce enriched fields + HubSpot
Matt Danese
Senior Demand Generation Manager. These stacks are built and used in production — not generated for a listicle.

Your CRM is only as useful as the data in it. Most inbound leads arrive with a name, an email, and maybe a company name — which tells you almost nothing about whether they're worth a rep's time or which segment they belong to. Without enrichment, your routing is a guess, your segmentation is incomplete, and your lead scoring is working off 40% of the picture. This stack automates the enrichment loop: pull firmographic, technographic, and contact-level data from Clay, Clearbit, and Apollo, run it through Claude for ICP scoring and segmentation, and push enriched records back to Salesforce and HubSpot ready for routing.

The Stack

Input
Clay Clearbit Apollo
AI
Claude
Output
Salesforce enriched fields HubSpot records

The Prompt

This stack is built around the Bad Lead Pattern Detection Prompt. Here's the abbreviated version — the full prompt with all variables and usage notes is on its own page.

Claude Prompt — Abbreviated
You are a B2B demand gen analyst scoring enriched leads against an ICP definition.

Review the enriched lead records below.
For each lead, evaluate: company size, industry, job title, tech stack, and geography.
Score each lead as Tier 1 (strong ICP fit), Tier 2 (marginal fit), or Tier 3 (disqualified).
For Tier 3 leads, identify the primary disqualifying signal.
For Tier 1 leads, note the strongest ICP signals to inform routing and messaging.
[ ... continued — see full prompt ]

The Workflow

  1. Set up Clay enrichment triggered by new CRM leads

    Configure a Clay table to trigger on new lead creation in Salesforce or HubSpot. Set Clay to automatically pull Clearbit firmographics and Apollo contact data for each record.

  2. Review enriched data before scoring

    Scan the enriched output for obvious gaps — Clay sometimes returns partial data on smaller companies. Fill critical missing fields manually before the scoring step.

  3. Paste enriched lead batch into the Bad Lead Pattern Detection prompt

    Include company size, industry, job title, tech stack, and geography for each lead. Add your ICP definition as prompt context so Claude scores against actual criteria, not generic assumptions.

  4. Review Claude's ICP tier assignments

    Claude assigns each lead a tier (Tier 1/2/3) with the primary signal driving the score. Review Tier 3 disqualifications before accepting them — data gaps can look like disqualifiers.

  5. Push enriched fields and scores back to CRM

    Update Salesforce and HubSpot with the new fields and tier scores. Use them to drive routing logic, segmentation lists, and campaign targeting for the next send.

What This Replaces

Related Stacks

New stacks drop weekly.

Each one includes the tools, the Claude prompt, and the workflow logic. Free — built for in-house B2B demand gen managers.

← Back to Stack Library

© 2026 The Demand Engineer · thedemandengineer.com