Visit statistics without cookies.

Why count at all.

Having a site and not knowing what happens on it is like running a konoba and not knowing how many tables were taken. You do not need every detail. You need a few answers: do people come, where from, what do they look at, and do they get in touch in the end. From that you can see whether an ad pays off, which page does its job and which one needs fixing.

Imagine a rental host whose statistics show that in spring a good share of visitors arrive from Austria and Germany through Google, while the site exists only in Croatian and English. That is a clear sign the next step is a German version, not another ad. Without the figures, that decision would be a guess.

What you see every month.

• Visits and visitors. How many people came and how many pages they looked at, day by day, so you can see when it is busy and when it is quiet.
• Where they come from. From Google, from Instagram or Facebook, from a message in which someone shared the link, or directly, by typing the address.
• Arrivals from AI tools. When someone gets a link to your site in ChatGPT or a similar tool and opens it, that visit usually shows up as coming from that tool.
• Which pages they read. Which page is read most, where people stay and where they leave.
• Which page brings messages. Every tap on the WhatsApp, call or e-mail button is counted against the page where it happened. So you can see which page not only draws people in but also gets them to make contact.
• Country and device. Which countries your visitors come from and whether they look on a phone or a computer. For a rental host it shows which languages the site most needs to get right.

The figures have limits too, and it is better to know them. I count the tap on the button, not whether the message was actually sent, because what happens inside WhatsApp cannot be seen from the site. Nor can it be seen when someone reads your name in an AI answer and then types it into Google. That visit looks like a visit from Google. Statistics show a direction, not every single person.

That is why it is most useful to watch changes rather than single figures. The number of visits in May says little on its own. The same number compared with April, after the site got a new language or after an ad started, says a great deal more.

How it is counted without cookies.

The usual statistics tools store a cookie on the visitor's device, a small file with a tag that lets them recognise the visitor next time, and often on other sites as well. That is why sites that use them must ask for consent first, and the visitor is greeted by a banner across half the screen.

The statistics I set up work differently. They store nothing on the visitor's device and do not follow anyone from one site to another across the internet. The visit is counted on the server, and tomorrow the same visitor is not recognised as yesterday's. No IP addresses, names or e-mail addresses are stored. You get totals, not a list of people.

That is why, for statistics of this kind, the site needs no cookie consent banner. A visitor opens the site and sees your business straight away, not a question about cookies. There is no extra script for a banner either, so the site stays lighter. On sorich.hr, the Legal page also says who processes that data.

Ad tracking tools that use cookies are a different matter. If a site ever gets them, it gets a consent banner too. The statistics I am writing about here work without them.

Where the statistics are used.

Visit statistics are part of the care plan for every website I build, for €150 a year, together with the domain, hosting, small changes within 24 hours and monitoring. Alongside Google and Meta ads they show what really matters: not how many people clicked the ad, but how many got in touch afterwards.

The same statistics run on sorich.hr, so what I offer is something I use myself. And if you wonder how many visits AI tools send you and why those tools pick whom they pick, the article on how AI decides whom to recommend covers it.