Warming up
Search by company domain, by a person's name and company, or by a LinkedIn URL. One at a time or a few thousand at once — each result carries a verification status and a confidence score.
Give it a domain and it finds the published or most probable contact address for the company. Give it a name and a company and it resolves that specific person. Paste a LinkedIn URL and it works from the profile. All three land in the same result table with the same status and confidence, so a mixed workflow does not mean a mixed set of tools.
Finding a plausible address and confirming a deliverable one are different operations, and a finder that only does the first hands you a bounce rate. Each result is verified as part of the lookup and returned with its status — verified, accept-all, or unverifiable — so the row already tells you whether it is safe to send to.
Results can be added directly to a campaign, exported as CSV, or kept in the Data Library where every bulk job across every tool is stored and re-openable. A finder whose output has to be manually re-imported somewhere else is doing half a job.
The email finder answers a narrow question precisely: what is the address for this company, or this person at this company. It exists next to the lead database rather than inside it because the two solve different problems — one builds a segment, the other resolves a specific target you already care about.
There are two routes to an address and a good finder uses both. The first is discovery: the address is published somewhere — a contact page, an imprint, a team page, a press release — and the job is to find and extract it. This produces the highest-confidence results, because the company itself put the address there.
The second is pattern inference. Most organisations use a consistent format — first.last, first initial plus last, first only — and once the pattern for a domain is known, a name can be resolved into a candidate address. This is probabilistic, which is exactly why the result carries a confidence score rather than being presented as fact.
It is an estimate of how likely the address is to be both correct and deliverable, and it should change what you do. A high-confidence, verified result can go into a campaign without further thought. A middling one is worth a second look, particularly if the domain is a catch-all. A low one should not enter a sending list at all.
The single most useful habit is to set a floor and enforce it. Sending to everything a finder returns, regardless of score, converts a useful tool into a bounce generator — and bounce rate is charged directly against the sending domain's reputation.
A catch-all server accepts mail for every address at the domain, including ones that do not exist, so verification cannot confirm the mailbox is real. The result is not wrong — it is unresolvable. Treat accept-all as a maybe, and keep it out of high-volume sends.
Upload a CSV of domains, or of names with their companies, and the whole file is processed. The job appears in the Data Library while it runs and stays there afterwards, so a two-thousand-row lookup does not depend on keeping a browser tab open, and the result set can be re-opened, re-exported or pushed to a campaign later.
Practical advice: clean the input first. Duplicate domains, redirect chains and misspelled company names all consume lookups and return nothing. Ten minutes de-duplicating a file routinely saves more allowance than any setting in the tool.
Use the lead database when you are building a segment: the shape of the request is a description of a kind of company, and you want everyone who matches. Use the finder when you already have a specific target — an account list from your CRM, a list of companies from an event, a set of profiles you shortlisted by hand — and what you are missing is the address.
They chain naturally in one direction. A database segment gives you companies; the finder resolves named people inside the ones worth approaching personally. The reverse rarely makes sense.
Addresses like info@, hello@ and contact@ are easy to find because they are published, and that is precisely what makes them weak targets. They land in a shared inbox, are frequently filtered aggressively, and are read by someone whose job is to triage rather than to buy.
They are not useless. For small businesses — a dental practice, a local agency, a trades firm — the published address is very often read by the owner, and it is the only address there is. For anything mid-market and up, resolving a named person is worth several times the effort.
In the UK and EU, a published business contact address at a corporate body sits on considerably firmer ground for B2B outreach than a personal address does. Inferring an individual's address by pattern is a step further and is treated as personal data. You still need a lawful basis, an honest identification of yourself, and a working opt-out in every message.
None of that is legal advice, and the analysis differs by jurisdiction. What Warmerly does is make the compliance mechanics automatic — unsubscribe links that work immediately, workspace-wide suppression, and a hard block on sending without a postal address on file.
Accuracy varies by how much a company publishes. Published addresses are near-certain; pattern-inferred addresses depend on how consistent the domain's format is. Every result carries a confidence score for exactly this reason — use it as a threshold, not as decoration.
Yes. Paste the profile URL and the lookup works from it, returning the same status and confidence as any other route.
The receiving server accepts mail for every address at that domain, so it is impossible to confirm the specific mailbox exists. Treat it as unresolved rather than valid, and keep those addresses out of high-volume campaigns.
Results are verified as part of the lookup. If a list has been sitting for weeks before you send, re-verify — addresses go stale, and a stale list is how bounce rates climb.
Into the Data Library, where every bulk job from every tool is stored. From there you can re-open it, export it, or push it into a campaign.
Domain, person or LinkedIn URL — one at a time or a few thousand at once, each result verified and scored.