The short answer
Company identification works by resolving the network connection a visit comes from to the organisation behind it. That works when the connection belongs to a company — an office network, a corporate uplink, a business ISP range. It does not work when the connection belongs to a household or a mobile carrier, because there is no corporate identity in the signal to attribute.
For B2B websites, roughly 20% to 45% of sessions typically resolve to a company. Pages with high purchase intent — pricing, product, documentation — usually sit at the top of that range; blog and top-of-funnel traffic sits lower.
Which visits stay anonymous
Four groups account for most of the unidentified share. None of them is a data-quality defect; they are properties of how people connect to the internet.
- Home offices and private broadband lines — see the separate answer on remote visitors.
- Mobile networks, where the carrier owns the address range rather than the visitor's employer.
- Consumer VPNs and privacy proxies, which deliberately hide the originating network.
- Small companies whose office traffic runs over an ordinary consumer contract.
Why we don't close the gap with guesswork
A vendor can raise its headline rate in two ways: better data, or looser matching. Looser matching means attributing a household connection to whichever company seems plausible, which produces sales lists full of companies that never visited you. The alternative — recognising the individual device across sites — is exactly the kind of person-level tracking that makes visitor identification a compliance problem in the first place.
lead.box does neither. When the connection cannot be attributed to an organisation, the visit stays anonymous. You still see it in your traffic totals; it simply carries no company badge. Any claim of near-total identification should make you ask which of the two shortcuts is behind it.
How to read your own rate
Your rate depends on your audience, not on a setting. A tool sold to enterprise IT teams sees a lot of office traffic and a high rate; a product bought by freelancers sees mostly private connections and a low one. Traffic source matters too: paid social often arrives on mobile, while direct and organic visits to documentation pages tend to come from desks.
lead.box shows the identification rate per page so you can size opportunities honestly instead of guessing. If the rate on your highest-intent pages is well below the typical range, the useful questions are usually about audience and channel mix — and that is a conversation worth having during the trial rather than after a purchase.
Related questions
Measure your own rate first
The only rate that matters is yours. The 14-day trial needs no credit card: install the snippet, let normal traffic run for a week and read the per-page identification rate.