127,560,445 business contact records across two corpora. This page lists every field you can filter on and export, explains how verification works, and is honest about what freshness means for a compiled dataset.
You search one corpus at a time. They overlap in the fields that matter for outreach and differ in what they are organised around.
Two normalised fields do most of the segmentation work. Email domain type tells you what kind of address you are about to unlock. Title level tells you how senior the person is, regardless of how their employer words the job.
Available on both corpora, filterable before you spend anything.
Company-owned domains. The bulk of the corpus and the right pick for B2B outreach.
.edu addresses at schools, colleges and universities.
Webmail providers (Gmail, Yahoo, Outlook and similar) attached to a business record.
Federal, state and municipal domains.
LinkedIn corpus only. Titles in the US corpus are free text and are searched as text.
CEO, CFO, CTO, COO, founders and owners
Vice presidents and senior vice presidents
Directors and heads of department
Managers, team leads and supervisors
Records that do not map cleanly to one of these bands are filed as Other and are still fully searchable by title text.
Not every record carries every field. These are approximate fill rates across the corpus, published so you can plan a campaign around what is actually there.
Because search is free, you never have to take these numbers on trust for your particular segment. Filter to the exact criteria you care about, add the verification-status or phone filter, and the result count tells you how many records satisfy it before you spend a credit.
When you spend a credit on a contact, the email address is submitted to a live verification service before the response returns to you. The result is one of a small set of statuses:
Results are written back to the source record, so the status you produced is the status the next search sees, for you and for everyone else. Roughly 14,400,000 US rows already carry a stored valid or cross-verified status from earlier checks, and you can filter on that status before you unlock anything.
This is a compiled dataset, not a live crawl. Records were assembled from public business filings and directories, company websites, and licensed and partner-sourced business data, and they carry the state of the world at the time they were compiled. People change jobs; companies close; phone numbers get reassigned.
What we do about that is deliberately narrow and verifiable: verification runs at unlock, and the result is written back to the source. So the field you are paying for is checked at the moment of purchase even when the row around it is older, and every unlock anyone performs improves the corpus for the next search.
What we do not claim: a headline accuracy percentage, real-time intent signals, or continuous re-verification of records nobody has asked for. If an address fails at unlock, you are not charged for it. That is the guarantee, and it is the only one we make.
Suppression is applied everywhere. Addresses submitted through Do Not Sell or Share stop being returned by search, unlock, export and the API.
Counted directly against the corpus. Territories and records with no usable state code are excluded from the table.
The state breakdown is a grouped count over the whole corpus, so it is computed in the background and cached rather than run on every page view. It is not available at this moment. The totals above are unaffected, and searching by state works normally — run a search with a state filter and the result count gives you the same number, live.
Searching costs nothing. Build the filter set you would actually use, look at the result count, and unlock a handful before you decide whether the data is right for you.