Lead Enrichment Pipeline

Upload a CSV of Leads, Get Back a Fully Enriched, ICP-Scored List

Raw lead lists are just names and emails. Before your team can prioritize outreach, someone has to research each company — size, industry, funding, tech stack — and decide whether it fits your ICP. The Lead Enrichment Pipeline automates that research, scoring each lead against your Ideal Customer Profile and writing enriched data back to your CRM.

What It Does

Trigger this workflow with a CSV file of leads (columns: first_name, last_name, company, email, title). Here's what happens:

1. CSV Parsing: The workflow reads your CSV and processes leads in batches. Deduplicates by company so research is done once per company, not once per contact.

2. Company Research: For each unique company, the workflow runs three web searches:

• Company size and headcount (current estimates)
• Tech stack (from LinkedIn, G2, and job listings which reveal what tools they use)
• Recent news and announcements (last 90 days)

3. AI Data Extraction: Claude processes the search results and extracts structured company data:

• Company size bucket (1–10, 11–50, 51–200, 201–1,000, 1,000+)
• Industry vertical
• Funding stage (bootstrap, Seed, Series A–C, public, etc.)
• Key technologies in their stack
• Recent news relevant to your product (fundraise, expansion, new hire signals)

4. ICP Scoring: Each lead is scored against your configured ICP criteria:

• Company size range (configurable min/max)
• Target industries
• Target job seniority (Director, VP, C-Suite, Head of)
• Funding stages you target
• Score: high / medium / low fit with a written reason

5. CRM Update: Enriched records are written back to your CRM (HubSpot, Salesforce, Pipedrive) via the CRM skill. All enrichment fields are tagged as "AI-enriched" with a timestamp.

6. Export: A complete enriched CSV is also exported with all original fields plus the enrichment columns, ready for import or filtering.

Who This Is For

Sales and marketing teams that import lead lists from events, LinkedIn Sales Navigator, ZoomInfo exports, or content marketing. Any team that receives raw leads and manually researches them before outreach benefits immediately from this workflow.

What's Included

• lead-enrichment-pipeline.yml — complete workflow with CSV parsing, multi-search company research, AI data extraction, ICP scoring, CRM update, and enriched CSV export
• Configurable ICP criteria (company size, industries, seniority, funding stage)
• Webhook endpoint for real-time enrichment (/webhooks/enrich-leads)
• Deduplication logic (research each company once)

Setup Steps

1. Install the workflow — copy lead-enrichment-pipeline.yml into your OpenClaw workspace
2. Configure your ICP — edit the icp_config section: target company size range, industries, seniority levels, and funding stages
3. Connect your CRM — authorize HubSpot or Salesforce in OpenClaw settings
4. Prepare your CSV — ensure columns are named: first_name, last_name, company, email, title
5. Run it — openclaw workflow run lead-enrichment-pipeline --csv-path ./leads.csv
6. Review the output — check the enriched CSV and CRM for a sample of records before running your full list

Understanding ICP Scores

• High fit: matches company size, industry, seniority, and funding stage criteria
• Medium fit: matches most criteria with one miss (e.g., right industry, slightly small company)
• Low fit: missing 2+ criteria — still worth contact if the specific reason doesn't matter for your product

Each score comes with a written reason (e.g., "High fit — 150-person Series B SaaS company, VP of Engineering matches buyer persona") so your reps can make judgment calls on borderline cases.

Why This Beats Manual Research

Manually enriching a 100-lead list takes 3–6 hours if done properly. In practice, most teams either skip it or do it superficially. With this workflow, a 100-lead list processes in under 10 minutes and produces richer data than most manual research would.

Result: reps start conversations with context, not cold. Pipeline quality improves because reps prioritize based on ICP fit, not arbitrary order.