How to track who visits your website, and what you cannot see

Search for how to track who visits your website and you will be sold three different things under one phrase. Two of them work. The third does not exist, and most of the pages selling it are careful never to say plainly what it returns.

Sorting them out first saves buying the wrong one.

The three things, and which are real

What aggregate analytics gives you

Sessions, pages, sources, devices, and the paths people take through the site. This is the boring answer and it is the one that changes decisions. Knowing that a third of your checkout starts come from one blog post is worth more than knowing the name of any single visitor, because it tells you what to do next.

Session recording sits alongside it. A replay shows the rage clicks on a control that is not a button, the form field that gets abandoned, the price nobody scrolls far enough to reach. That is not identification, it is observation, and it answers the question people usually mean when they ask who is visiting: what were they trying to do.

Why company lookup works less often than it is sold

The mechanism is simple. The visitor arrives with an IP address, and some IP ranges are registered to organisations rather than to internet providers. Match the range, name the company.

The failure cases are the majority of traffic.

• Anyone on home broadband resolves to their internet provider, not their employer.
• Anyone on a phone resolves to a mobile network.
• Anyone behind a VPN or a corporate proxy resolves to wherever that exits, which may be a different country.
• Small companies rarely hold their own IP ranges at all, so the ones you most want to name are the ones least likely to resolve.

What survives is a partial list of larger organisations on office networks. That can be useful for B2B outbound. It is not a record of who visited, and a tool that reports it as one is describing a sample as if it were a census.

Identifying a person needs the person

An anonymous visit carries an IP address, a user agent and whatever cookies you set. None of that is a name. A visitor becomes a known person at the moment they tell you who they are, by submitting a form, logging in, or clicking a link from an email you sent them. Everything before that is behaviour, not identity.

That is the practical reason lead capture beats visitor identification. The form is not a worse version of knowing who they are. It is the only version.

The consent part, which is not optional in the UK

Analytics and session recording both write to the visitor browser, which puts them in the class of non-essential cookies. In the UK that means a consent banner that actually gates the script rather than a notice that says cookies are in use while the script has already run.

The practical consequence is a measurement one. If a share of visitors decline, your analytics is a sample, and comparing a month before a banner to a month after it is comparing two different populations. Worth knowing before you conclude that traffic fell.

What to measure instead

1. Enquiries, by source. The number that pays wages, split by where the person came from.
2. The path into the enquiry. Which pages appear before a form gets submitted, not which pages get the most visits.
3. Where people stop. The step in checkout or the point on a long page where sessions end.
4. Returning against new. A rising share of returning visitors on a B2B site usually means a longer consideration cycle rather than a problem.

None of those needs a name attached to it. All of them change what you build next.