What Is Engagement score?

A summary of selected interactions

An engagement score assigns meaning to events the system can observe, such as forms, meetings or content interactions. The chosen events, weights, caps and time treatment determine the result. It is therefore a configured summary of behavior, not a direct measurement of a person's private intentions. A larger number can reflect more recorded events without establishing a stronger business need.

Keep event quality visible. Security systems and automated agents can generate opens or clicks, while some genuine interactions are not captured at all. Multiple events can also come from one action or be associated with an account rather than a known person. Do not let the apparent precision of a score hide those attribution and observation limits.

Separate interaction from audience fit

A student researching a topic may engage heavily without being a suitable buyer. A relevant executive may have little recorded activity but still have a real problem. Compare engagement with an independent fit assessment and current conversation evidence. A combined score can support a workflow, but its components should remain available so the seller understands why the record appeared.

Review time and duplication rules

Decide how older activity should affect current priority and how repeated events are counted. Follow the actual platform's supported behavior rather than assuming every system implements decay or deduplication identically. Test known events and inspect the resulting record before enabling a rule that changes ownership, stage or campaign eligibility. A correct event count does not prove that the chosen business action is appropriate.

Use engagement as evidence for a next question or prioritization choice, alongside suppression and relationship history. It should not override an explicit request to stop contact or automatically turn a contact into a qualified opportunity. The practical aim is an explainable queue informed by meaningful interactions, with uncertainty retained where the system cannot distinguish human interest from incidental or automated activity.