Useful analytics: measure decisions, not dashboards

The question is not, “How many events can we track?”
It is, “Which decision would we make differently if this number changed?”
Without that answer, data can become technical decoration: reassuring to look at, not very useful when a priority must be chosen.
Start with real objectives
A website can have very different goals: an enquiry, a quote, an application, an order, a document download, an event registration or the consultation of a decisive page. They do not all deserve the same weight.
Before opening Tag Manager or an analytics dashboard, define:
- which actions indicate genuine interest;
- which step in the journey is currently least clear;
- who will read the data and how often;
- which segments can genuinely change a decision;
- which information does not need collecting.
This also makes privacy simpler: restrained, proportionate and disclosed tracking is often more useful than an endless collection nobody can interpret.
Events are not conversions by default
A click, scroll or page view can be a useful signal. It does not automatically become a business result. A click on a phone number may deserve attention; a page view alone rarely explains why someone chose or abandoned a path.
Measurement therefore needs context: starting page, call to action, device, channel, journey stage and destination quality. Without turning every visitor into a record to follow everywhere.
On FacileAutoRicambi, tagging and advertising distinguished actions such as quotes and abandoned carts to make targeting more precise. At CGM Consulting, analytics and Tag Manager connected behaviour, segmentation and corporate/recruiting objectives. Different contexts, same criterion: measure behaviour only when it helps improve a concrete choice.
Search Console and analytics answer different questions
Search Console helps explain how a website is found: queries, impressions, clicks, pages and indexing issues. Analytics and tracking describe what happens after arrival: which pages are travelled through, which actions occur, and where a flow stops.
Mixing the two levels produces hurried conclusions. More impressions do not mean more enquiries. More visits do not prove that a page is clear. Even a high event count can indicate a problem if people click repeatedly because they cannot find the way forward.
The IvoryLab Guitar Shop case study reports a doubling of Google impressions and clicks within a few weeks during an initial work cycle. It is a useful organic-visibility signal, not a shortcut to claiming revenue or conversions that are not documented.
A dashboard should help choose
The best monthly reading is often brief: what changed, which hypothesis deserves a test, and which problem should be left alone because data does not yet support it.
For on-site SEO and organic strategy and systems that need tracking, I work on structure, events, data quality and interpretation. Not to produce charts for celebration, but to make the next intervention less dependent on instinct.