Why So Many Teams Still Think in Sessions and Pageviews
Years after the shift to event-based measurement models became standard, many analytics teams still think — and report — in the language of the previous paradigm: pageviews, sessions, bounce rate. This is not because event-based measurement is poorly understood technically. It is because the conceptual shift it requires is bigger than the technical migration that delivered it.
Session-based thinking organises data around a unit — the session — that is convenient but increasingly disconnected from how users actually interact with digital products. A single visit might involve watching a video, scrolling through content, opening a chat widget, downloading a resource, and leaving without converting, only to return three days later through a different device to complete a purchase. A session-based model fragments this into disconnected, awkward units. An event-based model treats each of these as a meaningful signal in its own right, connected to a user rather than confined to a visit.
What Changes When You Design for Events, Not Pages
Designing a measurement plan around events rather than pages starts with a different question. Instead of asking ‘which pages do we need to track?’ the question becomes ‘which actions represent meaningful progress toward our business objectives, regardless of where they happen?’ This reframing often reveals that the most important user actions are not page transitions at all — they are interactions within a page: video engagement, form field completions, content expansion, configurator usage, document downloads.
It also changes how you think about funnels. A page-based funnel assumes a linear sequence of page visits. An event-based funnel can represent the actual, often non-linear, paths users take — and can be rebuilt and re-segmented without needing new tracking, because the underlying event data already captures the granularity needed.
Making the Shift Practical
The practical risk in event-based measurement is the opposite of the problem it solves: rather than too little data, teams can end up tracking everything, producing event volumes that are technically rich but analytically useless because nobody can find the signal in the noise.
The discipline that makes event-based measurement work is the same discipline that underpins good KPI design — start from the business questions you need to answer, and build the event taxonomy backward from those questions. A well-designed event taxonomy is opinionated. It reflects a view of what matters. Teams that build their event structure without that view end up with technically correct but practically unusable data.
At Digital Analytics Lab, our GA4 and tag management implementations are built around event taxonomies designed for the specific business questions our clients need to answer — not generic templates that happen to be technically valid.
