Definition:
Tracking is the process of collecting, connecting and reviewing data about interactions with a website, application or campaign. It is used to understand where a visit came from, which actions took place and whether they produced the outcome being measured.
Each interaction can be recorded as an event with additional information: a page view, click, video play, form submission, purchase or call. Tracking turns these signals into data that can be analysed, but it does not necessarily identify a person or provide a complete view of their behaviour by itself.
What tracking is used for
Tracking connects investment in media and communication with what happens afterwards. A campaign, for example, may record the source advertisement or link, the visit to a landing page, a form submission and a later conversion. This sequence helps compare channels, locate drop-off points and assess results.
Tracking supplies the raw material for web analytics, but it does not replace interpretation. Nor does it decide which interaction deserves credit for a sale: that depends on the attribution model and the rules defined by the organisation.
How information is collected
An implementation may combine several methods. The choice depends on the channel, the data required and the relevant technical and legal constraints:
- Tags, scripts and pixels. Code placed on a page or creative records events and sends them to a measurement tool. A pixel is one such resource, not a synonym for the entire tracking system.
- Identifiers and storage. A cookie, local storage or a campaign identifier can help connect interactions within the applicable limits.
- Client-side tracking. Client-side tracking runs in the browser or device and may be affected by blockers, browser restrictions or loading failures.
- Server-side tracking. Server-side tracking processes selected events in infrastructure controlled by the organisation. It improves technical control, but it does not remove privacy obligations or automatically correct poorly defined data.
Application SDKs, APIs, server logs, UTM codes, CRM systems and call-tracking platforms may also be involved. These methods complement one another; they do not all observe the same interactions.
Measurement plans and tools
Before choosing a tool, it is useful to define which decisions the data should support. A measurement plan sets out objectives, events, parameters, naming rules, owners and tests. A data layer can standardise information generated by the site, while a tag manager helps deploy measurement code in a controlled way.
Google Analytics 4 and Adobe Analytics are used to analyse traffic and events; Contentsquare and Microsoft Clarity focus on the browsing experience; advertising and call-tracking platforms cover signals specific to their channels. No tool guarantees correct measurement when events are duplicated, parameters are missing or the implementation has not been tested.
Privacy and data quality
Tracking data may be incomplete because of consent choices, blockers, cookie restrictions, device changes or the separation between online and offline activity. Figures from different tools do not therefore always match and should not be interpreted as an exact record of every person.
A responsible implementation limits collection to necessary information, documents its purpose, applies the appropriate consent or legal-basis mechanisms, and establishes access and retention controls. It also needs regular testing to detect missing or duplicated events and incorrect values. Collecting more data does not automatically mean measuring better.