A technical study on how B2B teams should build, structure, and scale their prospecting lists — from raw data layers to the Waterfall ICP.
Every B2B database mixes two kinds of data: fields the person actually wrote about themselves, and fields a vendor's algorithm inferred about them. Most targeting failures begin here — in teams treating inferred fields as primary signal.
Written by the person. No algorithm. Use as your primary filter logic.
job_titleFree-text field — exactly what the person typed.
headlineSelf-description — the richest targeting signal available.
Inferred by algorithm. Hidden error rates. Use only as broad exclusion gates.
seniority_level«Founding Partner» at a 3-person firm classified as «C-Level».
job_function«Head of Growth» bucketed as «Business Development».
A non-normalized field's coverage depends on the strength of your keyword strategy — not on how long your title list is.
Exact match on a single normalized title captures a tiny fraction of the segment you think you're targeting. Run People Search for Marketing Director and you'll find roughly 18% of the marketing leaders at your target companies. The other 82% sits behind every other title humans actually use.
The intuitive fix — widen the People Search keyword list — trades one problem for another. You discover more contacts, but a smaller share of them are actually decision-makers.
Instead of asking «who do I want to reach?», start by asking «which companies belong in my market?» — then go contact-hunting one company at a time.
Your account list is a reusable asset — stable for 12–24 months. Build it once, refresh it on your cadence.
Score and route every contact you find — native priority, zero manual sort. Same logic, repeated per company.
Contact data refreshes — the company list never expires. Bad emails get replaced. Account fit stays.
Top-down, one company at a time. If P1 has no match, cascade to P2, then P3 — until a relevant contact is found. The hierarchy is your routing logic, not just a wishlist.
One query captures «Head of Marketing & Growth», «Director of Brand Marketing», «VP Demand Generation» — without listing every variation. Keyword logic beats exact match 2–4× on coverage.
The exclusion list is more valuable than the inclusion list. Open inclusions wide — control quality through precise exclusions.
| Priority | Profile | Route to | Channel |
|---|---|---|---|
| P1–P2 | C-level / Director High-touch, strategic |
AE | Multichannel |
| P3–P4 | Manager / Champion Evangelism, bottom-up |
SDR | Email + LinkedIn |
| P5 | Department sweep Long-term nurture |
Marketing | Email sequence |
| P6 | Founder / CEO Qualification fallback |
SDR | Short email |
People Search bulk export. No account list, no scoring — manual sort every cycle.
Company list + Waterfall ICP. CRM routing. Persistent TAM asset built to last.
Intent signals + multi-team ABM orchestration. Dynamic TAM from real-time data.
The foundation — everything else builds on a durable company list.
Remove all seniority/job_function primary filters. Replace with job_title + headline keyword combos. Remove all exact-match strings.
Levels 4–5 (headline + dept sweep) are non-negotiable — they capture champions. Build the exclusion list before the first run.
P1–P2 → AE. P3–P4 → SDR. P5–P6 → nurture. Configure once, apply every cycle. Eliminates manual triage permanently.
DR below 80% means 20%+ of TAM is invisible. Optimise this before reply rate, quality score, or MQL volume.
Map the P3–P4 first name as a variable on the P1 row. One column in Clay. Immediate, measurable lift.
Targeting performance is decided upstream — at the architecture level, not later in the copy or the channel. Build the right data layer, the right waterfall, the right routing — and reply rates compound. Skip it, and no amount of clever copy will fix what's broken at the foundation.