Measuring without tracking people

Visitor counts are the least useful measure and the most quoted. What you want to know: how many people do what you want, via which pages they arrive, where they drop out, and what it costs you to win one.

So first define what a result is: a submitted form, a booked appointment, an order, a phone call. Without that definition every chart is a pretty picture.

Look at quality too: a page with lots of traffic and zero enquiries tells a different story than a page with little traffic and three enquiries a week.

There are analytics tools that build no personal profiles, share no data with ad networks and keep data inside the EU. With them you see visitor counts, entry pages, referring sources and which actions get completed.

What you do not get is following one person across multiple sites, and detailed advertising measurement. For those who do not advertise, that is no loss.

The upside: no consent bar means no visitors dismissing it, and therefore numbers about everyone instead of about the half who pressed 'agree'.

Pick a handful of numbers and look at them at fixed intervals, rather than staring at a dashboard daily. What is changing, and because of what?

Look for patterns you can act on: a form often started and rarely finished, a page drawing traffic on a question you do not answer, a step in checkout where people vanish.

And watch the inbox: the questions coming in are free research. Every question that returns three times deserves a place on the site.