Analytics

A metric is any number that describes something that happened, such as visits, clicks, sign-ups or time on page. A KPI, or key performance indicator, is a metric chosen because it is tied directly to a goal. Many metrics exist, but only a few deserve to be KPIs. Starting from the goal, and not from the reports that happen to be available, keeps the analysis focused.

It helps to separate metrics of reach, such as impressions and visits, from metrics of engagement and of action, such as comments, sign-ups or requests. Volume alone says little: a large number of visits may bring few useful actions, and a small audience may be very relevant. Reading metrics together, and over time, gives more context than any single figure on its own.

Conversion rate is the share of people who completed a desired action out of all those who had the chance to do it. The action, called a conversion, can be a purchase, a form submission, a newsletter sign-up or a download. Defining exactly what counts as a conversion, and who counts in the base, is the first step. Two reports only compare fairly when they use the same definition.

A rate always depends on its denominator. A page seen by few people can show a high rate that means little, while a busy page can show a lower rate and still produce more completed actions. It is also useful to look at the steps before the conversion, since drop-offs in a form or checkout reveal where people hesitate, without assuming the cause before testing.

UTM parameters are labels added to the end of a link, such as source, medium and campaign name. When someone clicks, the analysis tool reads these labels and records where the visit came from. Without them, many visits end up grouped as direct or unknown. A written naming convention, with consistent lowercase and no spaces, avoids duplicated entries in reports.

Attribution is the rule used to assign credit for a conversion to one or more touchpoints. A person may see an ad, read an email and later search for the site name before acting. Models such as first touch, last touch or shared credit tell different stories about the same path, so no model is neutral, and the choice should be stated whenever results are presented.

A metrics dashboard is a single screen that gathers the indicators a team checks regularly. A good one answers a few questions, shows the period compared and avoids decoration. Too many charts make decisions harder, so it is worth keeping separate views for daily monitoring, monthly review and deeper investigation, each with an owner who maintains it.

Collecting data also brings responsibilities. General good practice is to collect only what is needed, explain in plain language what is measured, ask for consent where appropriate, protect access and avoid storing personal details in reports. Browsers, devices and platforms change how tracking works, so some gaps in the data are normal and should be noted in the analysis.