Companies we crawled ourselves, filtered by what their website reveals

What 'crawled ourselves' actually buys you

Three things. The first is freshness in the places that matter: a company whose site went down, whose contact page changed, or whose stack moved is re-observed rather than remembered. The second is the filter set, technology detection, contact-route detection and site signals only exist if you have looked at the page.

The third is honesty about coverage. When you own the pipeline you know precisely how many records have a phone number, because you know how many pages had one. A reseller has to trust an upstream count, and the incentive runs the wrong way.

Contactable is a much smaller number than crawled

The count of companies in a database and the count you can actually email are very different numbers, and the gap is not small. A large majority of crawled company records carry firmographic data (sector, location, size, stack) with no published contact address anywhere on the site. Measured on the live database today: 9.7M+ of those records carry a contact route. That is an estimate, and it moves as the crawl runs.

Warmerly reports the contactable figure as the headline, not the crawl figure, and shows both. It is the less impressive number and it is the only one that describes what you can do. A vendor quoting a total record count while you are shopping for an email list is describing a different product to the one you are buying.

Suppression applies before you ever see a row

Two lists are enforced. Companies that have opted out of appearing in the database do not appear in anyone's results. And when you add leads to a campaign, your own workspace suppression list is applied, so an address you have already been asked never to contact cannot re-enter your outreach through a lead search.

That second one closes the most common route back to a suppressed contact. Suppression is normally thought of as something that happens at send time; applying it at import time as well is what stops the list from being quietly rebuilt.

Verify before you send, without exception

A published address on a website is not a guarantee of a working mailbox. Pages go stale, staff leave, catch-all domains accept everything and deliver nothing. Running an export through verification before it enters a campaign is the difference between a 1% bounce rate and a 9% one, and a 9% bounce rate is how a healthy sending domain ends up on a blocklist.

Verification is in the same product for exactly this reason. It is not an upsell bolted on to the data. It is the step between having a list and being allowed to send to it.

Building a segment worth sending to

1. Start with the signal that qualifies, not the geography. 'Runs Shopify and Klaviyo' is a better first filter than 'in the UK', because it says something about whether they need you.
2. Add the reachability channel you actually intend to use, so you are not counting companies you cannot contact.
3. Read the coverage bars. If the segment is thin on the field your personalisation depends on, narrow it or change the approach before exporting.
4. Layer location and size last, to size the list rather than to define it.
5. Export, verify, then import to a campaign, in that order, every time.

The common mistake is inverting steps one and four: building a large geographic list and then hunting for a message that fits it. A narrow, well-qualified segment with a message that could only have been written for it outperforms a list ten times its size, and costs less to send to.

What this data is not

It is company data with published contact routes, not a database of named individuals with personal email addresses. Where a person's address is needed, the email finder resolves a specific person at a specific company on demand rather than shipping a bulk list of individuals.

That distinction is partly a legal one. Published business contact details sit on much firmer ground under GDPR than a warehouse of personal addresses assembled from scraping, and it is a line worth being on the right side of.