Data enrichment strategy skill

Data Enrichment Strategy

Overview

Every B2B go-to-market team needs enrichment. No single provider covers everything, and the stack you choose determines your data quality, cost structure, and operational complexity for years.

This skill walks through provider selection, waterfall architecture design, build-vs-buy decisions, and ongoing data hygiene. The output is a complete enrichment architecture tailored to the user's ICP, budget, and volume.

When to Use

• "What enrichment providers should we use?"
• "Apollo vs ZoomInfo vs Clay — which is right for us?"
• "Design our data enrichment architecture"
• "How do we get 90%+ email coverage?"
• "Should we build or buy our enrichment pipeline?"

Authoritative Foundations

The enrichment provider landscape is well-documented by DAMA-DMBOK, Open Data Institute, Ziellab, and the GTM engineering community. The core design principle: no single provider covers more than 60-75% of B2B contacts. Waterfall enrichment — chaining multiple providers in sequence — is the standard pattern for hitting 85-92% coverage.

Three separate waterfalls (company, email, phone) is the Ziellab architecture: each field type uses different providers optimized for that data type, and the waterfalls run independently.

Prerequisites

• ICP defined (company size, industry, geography, buyer personas)
• Monthly enrichment volume estimate
• Budget range for data + platform costs

Step-by-Step Process

Phase 1: Intake

Ask the user:

1. Target segment (SMB, mid-market, enterprise, or all)?
2. Primary geography (US, EU, APAC, global)?
3. Monthly enrichment volume?
4. What data points are critical? (email, phone, firmographics, tech stack, intent)
5. Existing tools? (CRM, sequencer, Clay, etc.)

Phase 2: Provider Selection

Provider	Best For	Email Coverage	Phone Coverage	Firmographics	Cost Profile	Notes
Apollo	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
ZoomInfo	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Clay	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Clearbit (HubSpot)	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
People Data Labs	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Cognism	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Lusha	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Hunter	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Findymail	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights
Prospeo	Evaluate for the target segment	Measure	Measure	Verify fields	Check current quote	Review terms and data rights

Phase 3: Architecture Design

Stack by stage:

Startup / <$2M ARR, <5K contacts/month:

• Apollo (all-in-one: data + sequencing) = $59-99/month
• Add verification service = $50/month
• Total: $150/month

Growth / $2-10M ARR, 5-30K contacts/month:

• Clay ($149-800/mo) + Apollo data + 1-2 fallback providers
• Waterfall: LeadMagic Email Finder → Apollo → Hunter → verification
• Add CRM integration for automated enrichment
• Total: $800-1,800/month

Enterprise / $10M+ ARR, 30K+ contacts/month:

• ZoomInfo for firmographics + org charts
• Clay for multi-source orchestration and workflow
• Waterfall: ZoomInfo → LeadMagic → PDL → Cognism → Claygent
• n8n for custom pipeline logic
• Total: $3,000-8,000/month

Phase 4: Build vs Buy

Factor	Buy (SaaS platform)	Build (API-first)
Setup time	Hours to days	Weeks
Monthly cost	$150-1,800	$500-3,000+ (engineering)
Flexibility	Limited to platform capabilities	Unlimited
Maintenance	Platform handles	Your team handles
Best for	<30K contacts/month, standard workflows	>30K contacts/month, custom pipelines

Hybrid pattern: Clay for workflow + n8n/API pipeline for high-volume or custom enrichment. Clay handles the ad-hoc and mid-volume; API handles the scale.

Phase 5: Data Hygiene Cadence

Activity	Frequency
Email verification	Every 90 days (or before each campaign)
Firmographic refresh	Quarterly
Technographic snapshot	Monthly
CRM deduplication	Weekly
Full enrichment audit	Quarterly

Output Format

Enrichment strategy document:

# Enrichment Architecture for [Company]

## Current State
- Volume: X contacts/month
- Current providers: [list]
- Coverage: X%
- Cost: $X/month

## Recommended Stack
- Primary: [provider]
- Fallbacks: [providers in order]
- Verification: [provider]
- Orchestration: Clay / n8n / API

## Waterfall Design
- Email waterfall: [ordered list]
- Company waterfall: [ordered list]
- Phone waterfall: [ordered list]

## Cost Model
- Platform: $X/mo
- Provider credits: $X/mo (est. X contacts × $Y/contact)
- Verification: $X/mo
- Total: $X/mo

## Implementation Timeline
- Week 1: Provider setup
- Week 2: Pilot (500 contacts)
- Week 3: Validation and tuning
- Week 4: Production rollout

Quality Check

• Provider selection matched to ICP segment and geography
• Waterfall ordered by cost-per-hit, not just coverage
• Verification step included after every email finder
• Data hygiene cadence defined
• Cost model complete (platform + credits + verification)
• Build-vs-buy decision documented with rationale

Common Pitfalls

1. Single-provider lock-in. One provider covers 60-75% max. If ZoomInfo is your only source, you're leaving 25-40% of contacts unreachable.

2. Over-buying for the stage. $25K/year ZoomInfo contract for a $500K ARR startup is capital misallocation. Start with Apollo, add tools as volume and deal size justify them.

3. No verification in the architecture. Enrichment without verification means sending to stale emails. Build verification into the design, not bolted on later.

4. Ignoring data decay. Data you enriched 6 months ago is 12-18% stale. Design for recurring enrichment, not one-time projects.

5. Platform as system of record. Clay is a workspace, not a CRM. Push enriched data to your CRM and treat it as the source of truth.

Execution Artifacts

• references/framework-notes.md — Named frameworks and reference tables
• templates/output-template.md — Deliverable shell for agent output
• scripts/check-output.py — Lightweight deliverable validator

Related Skills

• lead-enrichment: Execute enrichment on a specific list
• waterfall-enrichment: Detailed waterfall implementation
• clay-automation: Clay-specific workflow design
• api-enrichment: API-first enrichment pipeline
• crm-integration: CRM setup for enriched data