IcyPeas email finder: it returns an address even when it found nothing. Real hit rates on three lists
IcyPeas email-search returns an email with certainty 'not_found' when it only guessed a pattern. Counting every returned address inflates the yield by about 30%. Measured hit rates: 51% on consumer brands, 16% on big consumer names, 55% on small professional firms.
| List | Searches | ultra_sure | probable | not_found (guess returned) | Empty | Usable hit rate |
|---|---|---|---|---|---|---|
| Consumer brands and creators, US | 512 | 261 | 20 | 92 | 139 | 55% |
| Large consumer brands (addresses already dead at a verifier) | 19 | 3 | 0 | 13 | 3 | 16% |
| Small accounting firms, domain checked against the company name | 90 | 50 (ultra_sure + probable) | 55% |
What drives the hit rate
The accuracy of the domain you pass. On small professional firms whose domain was matched to the company name beforehand, more than half of the searches returned a sure address. On large consumer brands, whose public pattern is well known but whose employees rotate, the tool mostly returns a pattern guess. Three sure finds out of 19 still passed a strict verifier (2 pure, 1 catch-all), so it is worth a last-resort call: low yield, almost free since guesses are not charged.
Rules
- Filter certainty in ('ultra_sure', 'probable') before using an address; send the rest back into the cascade.
- Place it last in the cascade: free discovery-database lookups first, then a LinkedIn-based lookup, then IcyPeas, with a daily cap.
- Never push a 'not_found' address to a verifier or a campaign; it is a guess with a verifier's cost attached.
What this measurement does not say
- Hit rates depend on the list; the three runs here are consumer-facing or small-firm lists. B2B SaaS mid-market lists were not measured.
- Certainty labels are the vendor's; we checked only 3 'ultra_sure' addresses against a verifier on the second run.
Questions
Why does IcyPeas return an email it did not find?
The result carries a certainty field; 'not_found' means a pattern guess was returned. On 512 searches, 92 results (18%) were such guesses.
Related measurements
- MoltSets LinkedIn-to-email: 46% hit rate on profiles the discovery database had no email for, but its grade does not predict bounces
- How much of a 'verified' B2B contact list is actually catch-all or dead: GetLeads, Apollo, LeadMagic measured
Want this run on your pipeline?
We operate multichannel outbound (cold email, LinkedIn, paid retargeting, newsletters) for B2B companies, and we measure it the way these pages do.
Book a call