Data
Apollo-first enrichment waterfall
Stage 1 Apollo, then Hunter, Dropcontact, Prospeo. 92% valid on 47k rows — not four providers in parallel.
The loop
- 01Clean the input before you spendFirst/last, company, domain, LinkedIn URL, title. Dedupe. Split US vs EU.
- 02Stage 1: Apollo on every row60–65% of matches. Hit with email → done. Else queue Stage 2.
- 03Stage 2: Hunter Email Finder+15–20%. Keep only confidence ≥70 and verification = valid.
- 04Stage 3 Dropcontact, then Stage 4 ProspeoDropcontact +8–12% (LinkedIn URL lifts match). Prospeo +5–8% on what’s left.
- 05Verification pass, then stopZeroBounce / NeverBounce / Bouncer. Drop disposable, role-based, catch-all.
GENESIS ran a four-stage waterfall on 47,000 EU/US prospects and published the math. Apollo-only left about 37% of a 5,000 UK SaaS test dark (63% coverage). Overlap across Apollo, Hunter, Dropcontact, and Prospeo is only ~40–45%, so a cascade fills gaps without paying four times. Parallel on 10k rows would be ~$5,200; the waterfall was $1,120 — $0.114 per valid email, 78% cheaper.
Route geography: US Apollo-first, EU Dropcontact-first. Clean the CSV before you spend. Hunter confidence ≥70 and verification = valid, or you are guessing. Then a verification pass (ZeroBounce / NeverBounce / Bouncer). They reported 92% valid coverage and 94%+ verification. A waterfall is a sequence with a stop condition, not a pile of APIs.