The Cut Call: Auditing a Clay Waterfall Enrichment Run
Objective: Given a real 5-provider Clay waterfall enrichment log (attempts, hits, cost per stage), compute blended coverage and cost-per-match, then decide which provider to cut from the chain.
You're the growth marketing analyst at Mailchimp validating a new B2B outbound motion for its email/SMS suite. Before scaling the list, you need to know whether the 5-provider waterfall is actually worth its added complexity over a single vendor.
Import the waterfall log, compute cost-per-successful-match at each stage, calculate blended coverage, and recommend cutting or keeping the lowest-ROI provider with the number to back it up.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, no account friction, handles the formula work for this audit without any paid tool
The process
1 step
Step 01 of 01
The lesson's waterfall deep-dive chains providers instead of relying on one: try Apollo first, then Hunter, then Clearbit, and so on, paying only on a successful hit, so coverage climbs from a typical single-provider 30% toward 80%+ without proportionally climbing cost.
On a 2,500-contact list run through 5 providers in sequence, blended coverage lands at 72.8% but one provider is barely moving the needle at a steep price. Which provider gets cut, and what does keeping it actually cost?
Procedure
- Import the 5-stage log: Apollo (2,500 attempted, 780 hits, $200 spend), Hunter (1,720 attempted, 430 hits, $86 spend), Clearbit (1,290 attempted, 360 hits, $154.80 spend), People Data Labs (930 attempted, 210 hits, $27.90 spend), Crunchbase (720 attempted, 40 hits, $108 spend)
- Add a formula column: cost-per-match = spend / hits for each stage
- Sum hits across all 5 stages and divide by 2,500 to get blended coverage
- Sum hits and spend for the 4 stages excluding Crunchbase, recompute blended coverage and cost without it
- Compare the marginal coverage lift from Crunchbase against its cost-per-match versus the 4-provider blended average
Provider Attempted Hits Spend Cost/Match Apollo 2,500 780 $200.00 $0.256 Hunter 1,720 430 $86.00 $0.200 Clearbit 1,290 360 $154.80 $0.430 PeopleDataLabs 930 210 $27.90 $0.133 Crunchbase 720 40 $108.00 $2.700 5-provider blended: 1,820/2,500 = 72.8% coverage, $0.314/match 4-provider (no Crunchbase): 1,780/2,500 = 71.2% coverage, $0.263/match
Healthy
Blended cost-per-match stays under roughly $0.40 while coverage clears 70%.
Unhealthy
A provider's marginal cost-per-match spikes 5-10x above the blended average while contributing under 3% of total coverage.
What this means
Order and count both matter — a 5th provider recovering only 40 more contacts at $2.70 apiece isn't paying for itself once you can see the marginal math.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Crunchbase stage adds a 1.6-point coverage lift at roughly 10x the blended cost-per-match | Drop Crunchbase from the waterfall and redeploy its $108 budget toward raising the People Data Labs attempt cap | 30 min |
Final deliverable
A revised 4-provider waterfall order with blended coverage and cost-per-match projections, plus a one-line recommendation on Crunchbase backed by the marginal-cost number.
See a reference example
Klaviyo, Q3 waterfall audit (excerpt) Provider order: Apollo -> Hunter -> Clearbit -> People Data Labs Blended coverage: 71.2% (1,780 / 2,500) Blended cost-per-match: $0.263 Recommendation: cut Crunchbase from the chain. It contributed 1.6 points of coverage at $2.70/match, roughly 10x the blended average. Redeploy its $108 budget toward a higher People Data Labs attempt cap instead.
Success criteria
You're done when you can:
- Correctly computes cost-per-match for each of the 5 provider stages
- Correctly computes blended coverage with and without the cut provider
- Recommends cutting or keeping the lowest-ROI provider with the marginal cost-per-match number to back it up