Find the address, with a confidence score you can act on

How finding actually works

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.

What the confidence score is telling you

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.

Bulk lookups

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.

When to use the finder and when to use the database

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.

Role addresses, and when to avoid them

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.

The legal position, briefly

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 launch check that warns you when your workspace has no postal address on file. Including a valid postal address in your commercial email stays your responsibility.