First of all, it means having the ability - understood as know-how, but also as the correct choice of tools - to have a complete (I was about to say holistic) vision of the data regarding our activity on the web.
By this I mean that a correct data measurement strategy is only one that does not look at individual phenomena in isolation from the context , but on the contrary tries to understand how each action of our users influences subsequent ones, and how it is the result, indicator, signal of a behavior determined by the context in which our user finds himself interacting.
Let me try to explain myself better.
When we are faced with data coming from our armenia phone number website, or perhaps resulting from our digital advertising campaign, our attitude is, in general, to observe the navigation flows as an indistinct whole.
We observe the data by usually focusing on the so-called clickstream : how many sessions are opened, how many pages are viewed, possibly (but we are already in a moderately advanced context) how many objectives are achieved and how many macro-conversions are generated.
If we are managing a campaign on Facebook to promote a sponsored post (perhaps to spread a blog article) we will take care to check the number of impressions, the average cost per click, the number of landings on our landing page and so on, continuing to observe the pipeline that takes the user along the conversion funnel, up to our goal.
What's wrong with this approach to web analytics? Everything and nothing, really.