AI Stack for CRM Data Hygiene

Salesforce/HubSpot export + Clearbit → Claude → Deduped + enriched CRM records + Slack audit report
Matt Danese
Senior Demand Generation Manager. These stacks are built and used in production — not generated for a listicle.

CRM data degrades faster than most teams realize. Contacts change jobs. Companies get acquired. SDRs enter records inconsistently for three years until the database is a patchwork of contradictory information. The result is segmentation you can't trust, routing logic that fires on bad data, and attribution reports built on a foundation that doesn't reflect reality. This stack runs a systematic hygiene pass: export CRM records, enrich with Clearbit for current firmographic data, run through Claude for duplicate detection and stale record flagging, and get a ranked cleanup list you can act on this week.

The Stack

Input
Salesforce/HubSpot export Clearbit enrichment
AI
Claude
Output
Cleaned CRM records Slack audit report

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 marketing ops analyst running a CRM data hygiene audit.

Review the CRM records and Clearbit enrichment comparison below.
Flag records where: job title has changed, company domain doesn't match, email is likely
invalid, or the record appears to be a duplicate of another in the dataset.
Rank your findings by severity: critical (actively breaking routing or scoring),
moderate (degrading segmentation), and low (cosmetic cleanup).
[ ... continued — see full prompt ]

The Workflow

  1. Export CRM records for hygiene review

    Pull all contacts and leads created or modified in the last 90 days from Salesforce or HubSpot. Include key fields: email, company, job title, lead source, and last activity date.

  2. Run the export through Clearbit

    Send the record list to Clearbit to check for job changes, company accuracy, and email validity. The delta between your CRM data and Clearbit's current data is your hygiene problem list.

  3. Paste the comparison into the Bad Lead Pattern Detection prompt

    Include both your CRM data and the Clearbit enrichment output side by side. Claude compares the two, identifies mismatches, and flags records that should be updated, merged, or removed.

  4. Review Claude's ranked audit findings

    Claude produces a severity-ranked cleanup list: critical issues breaking routing or scoring, moderate issues degrading segmentation, and low-priority cosmetic fixes. Start with critical.

  5. Execute cleanup and post audit report to Slack

    Process the cleanup list in Salesforce or HubSpot. Post the audit summary to your marketing ops Slack channel so the team knows what changed and why. Schedule the next audit for 90 days out.

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