list-building-framework-md.md
B2B List Building
List building isn't about getting the most contacts — it's about getting the right contacts with the right data to enable personalized, relevant outreach. Quality > Quantity, always.
The 3 Pillars
- Coverage — Cast a wide net (multi-source)
- Enrichment — Add context & intelligence
- Precision — Filter & prioritize
The 8-Phase Framework
| Phase | What You Do | Key Outcome |
|---|---|---|
| 1. Define ICP | Firmographics, technographics, signals, psychographics | Targeting criteria |
| 2. Company Discovery | Multi-source search (Clay + Apollo + LinkedIn + more) | Raw company list (80–90% TAM coverage) |
| 3. Company Enrichment & Scoring | Validate firmographics, add signals, tier A/B/C | Prioritized company list |
| 4. People Discovery | Multi-source people search at tiered companies | Raw contact list |
| 5. Contact Enrichment & Validation | Email waterfalls, phone waterfalls, LinkedIn URLs | Verified contact data |
| 6. Deduplication | Dedupe at company level (domain) AND people level (email/name) | Clean, no-duplicate list |
| 7. Personalization & Segmentation | Segment by industry/size/persona/signal, add AI icebreakers | Campaign-ready segments |
| 8. Activation | Push to CRM, sequencer, or export | Outreach launched |
For detailed walkthroughs of each phase, see list-building-deep-dives.md. For industry-specific directories to scrape, see list-building-directories.md. For the full 62 data sources table, see list-building-data-sources.md.
Golden Rules
- Multi-source everything. One provider = 50–60% coverage. Two providers = 80–90%. The math is simple.
- Enrich before you reach out. Raw lists convert at 1–2%. Enriched, scored, personalized lists convert at 5–15%.
- Dedupe religiously. Duplicates kill deliverability and credibility. Always dedupe at company AND people level.
- Validate emails. 30–40% of emails in databases are invalid. Validation saves your sender reputation.
- Personalization at scale. Use AI to personalize, but always based on real data points (not generic fluff).
- Test & iterate. Track which sources give best data, which enrichments correlate with conversions, which segments perform best.
Phase 1: ICP (Ideal Customer Profile)
Why It Matters
- Right-fit companies convert 3–5x higher than spray-and-pray
- Shortens sales cycles (they have the pain, budget, authority)
- Improves retention (ICP-fit customers churn less)
"Our product works for everyone!" = the mistake most beginners make.
The 4-Layer Framework
| Layer | What It Covers | Example Criteria |
|---|---|---|
| Firmographics | Industry, size, revenue, location, funding stage | B2B SaaS, 100–500 employees, $10M–$50M ARR, Series A-B |
| Technographics | Tech stack, tools, platforms | Uses Salesforce + Outreach/SalesLoft, has marketing automation |
| Behavioral Signals | Hiring, funding, expansion, leadership changes | Recently raised Series B, hiring RevOps, expanding to EU |
| Psychographics | Growth mindset, innovation vs stability, data-driven culture | Hypergrowth, early adopter, customer-centric |
ICP Scoring System
Build a formula column in Clay:
| Criteria | Points |
|---|---|
| Right industry | +20 |
| Right company size | +20 |
| Right revenue range | +15 |
| Uses key tech | +15 |
| Recent funding | +10 |
| Hiring for relevant role | +10 |
| Target location | +10 |
Tiers:
- A-tier (80–100): Perfect fit, prioritize
- B-tier (60–79): Good fit, pursue
- C-tier (40–59): Okay fit, nurture
- D-tier (<40): Poor fit, disqualify
ICP Document Template
Must-Haves (Non-negotiable):
- Industry: [X, Y, Z]
- Company size: [X–Y employees]
- Revenue: [$X–$Y]
- Location: [Regions]
- Tech stack: [Must use X tool]
Nice-to-Haves (Bonus):
- Funding stage, hiring signals, complementary tools
Disqualifiers (Anti-ICP):
- Too small, wrong industry, wrong tech stack
For the full ICP deep dive (how to build, examples, common mistakes), see list-building-deep-dives.md.
Phase 2: Multi-Source Company Discovery
The Coverage Problem
| Strategy | Coverage | Companies Found (out of 10K TAM) |
|---|---|---|
| Single source | ~60% | 6,000 |
| Two sources | ~85% | 8,500 |
| Three sources | ~92% | 9,200 |
Two sources = 42% more companies than one.
Source Selection Matrix
| Your ICP | Primary | Secondary | Tertiary |
|---|---|---|---|
| Tech/SaaS (US/EU) | Clay Find Companies | Apollo | LinkedIn Sales Nav |
| Enterprise (Fortune 500) | ZoomInfo | LinkedIn Sales Nav | Apollo |
| Startups (funded) | Crunchbase | Clay Find Companies | LinkedIn Sales Nav |
| International (Asia, LATAM) | Apollo | LinkedIn Sales Nav | Local directories |
| Non-tech (Manufacturing, Retail) | Apollo | LinkedIn Sales Nav | Industry directories |
| Niche markets | Web scraping/Claygent | Apollo | Industry associations |
Primary Sources
Clay Find Companies
- Best for: Tech/SaaS, US/EU, technographic filtering
- 60M+ companies, built into Clay
- Weakness: weaker in Asia/LATAM, traditional/offline businesses
Apollo
- Best for: Broad coverage, international, non-tech industries
- 250M+ contacts, 60M+ companies
- Export: CSV → import to Clay
- Free tier: 50 exports/month
LinkedIn Sales Navigator
- Best for: Most current data, growth signals (headcount growth, hiring, recent posts)
- Unique filters not available elsewhere
- Export via Phantombuster, Evaboot, or manually
- $99–$149/month
Specialized Sources
- ZoomInfo — enterprise, high-accuracy ($15K–$30K/year)
- Crunchbase — funded startups, recent funding rounds ($29–$99/month)
- Web Scraping / Claygent — niche directories, association lists, conference attendees
- Your CRM — closed-lost (6+ months = gold), churned, upsell
Alternative Discovery Methods
Local Prospecting (when targets don't show up in B2B databases):
- Openmart — US-based local businesses
- Google Maps scraping via Clay — any niche with physical presence
- Import → Clay → enrich
Directory Scraping:
- Use Instant Data Scraper (Chrome extension)
- Export CSV → import to Clay → enrich
- See list-building-directories.md for industry-specific sources
TAM Scoping via DiscoLike:
- Discovers lookalike websites based on keywords/tech tags
- Not limited to LinkedIn data
- Good when Apollo/Clay/LinkedIn fail
- ~1 credit per row to qualify
- Filter by country, language, similarity variance, company start date
Phase 3: Company Enrichment & Scoring
Must-have enrichments:
- Firmographics validation (employee count, revenue, location)
- Technographics (tech stack = shows budget and sophistication)
- Funding/financials (recent funding = budget to spend)
- Hiring signals (hiring for roles related to your solution = active need)
- News/triggers (expansion, leadership changes, product launches)
Then tier/score: A-tier → B-tier → C-tier using formulas or AI in Clay.
Phase 4: People Discovery
Multi-source again:
- Apollo People Search — good coverage, job title filtering
- Clay's Find People — different data set, finds people Apollo misses (and vice versa)
- LinkedIn scraping — most current, slower/more expensive, harder to scrape over time
- Waterfall approach — try multiple sources per company to maximize contact discovery
Phase 5: Contact Enrichment & Validation
Email waterfalls (3–5 providers):
- LeadMagic, Prospeo, Findymail, etc.
- Personal emails first, catch-all/generic as fallback
- Validation to remove bounces
Phone waterfalls (if cold calling):
- Direct dials, mobile numbers
- Multiple providers for coverage
Personalization data:
- Recent job changes, education, location, interests
- Company news, recent posts, mutual connections
See Clay's data provider tests for provider performance by region.
Phase 6: Deduplication (Critical)
Dedupe at two levels:
- Company level — dedupe by domain before finding people
- People level — dedupe by email (or full name to save credits, accepting small risk of false deduplication)
Why: Duplicate outreach = spam = blacklisted domain. Never skip this.
Sequencer safeguard: Most tools (Instantly, Smartlead, HeyReach, Lemlist) have "Skip lead if in Workspace" toggle — use it as a safety net, not a replacement for proper deduplication.
Phase 7: Personalization & Segmentation
Segment by:
- Industry (different messaging)
- Company size (different pain points)
- Persona (different value props)
- Trigger/signal (different hooks)
Add personalization columns:
- AI-generated icebreakers referencing recent news, posts, achievements
- Custom first lines based on tech stack, hiring signals
- Relevant case studies/social proof from their industry
Phase 8: Activation
Options:
- CRM sync — push to Salesforce, HubSpot (with deduplication)
- Sequencer — Instantly, Smartlead, Lemlist, Outreach
- Manual export — CSV for one-off campaigns
- Clay campaigns — send directly from Clay
Conditional logic: Only push A-tier to expensive sequences, B-tier to nurture, C-tier to content campaigns.
Common Mistakes
- Using only one data source (leaving 40% of TAM on the table)
- Skipping enrichment (raw data = low conversion)
- No deduplication (kills deliverability)
- Generic personalization ("I saw you work in tech" is not personalization)
- Not validating emails (bounces destroy sender reputation)
- Treating all leads equally (tier and prioritize)
- Building lists without a clear ICP (garbage in = garbage out)
- Static ICP (revisit quarterly based on what actually converts)
- Ignoring anti-ICP (knowing who NOT to target is equally important)
Key Principle
List building is not a one-time task — it's an ongoing system. The best teams:
- Refresh lists monthly (data decays ~30% per year)
- Continuously test new data sources
- Refine ICP based on what actually converts
- Build feedback loops from sales to improve targeting
Built by Frontal & Ivan Falco. For questions on implementation or anything not covered here, reach out to Ivan directly on LinkedIn.