Studio Pulse

Visitors: exact distinct analytics IDs, not verified people. Basic measurement is cookieless, so identity can change between visits or days. Returning-person counts cannot be reliably inferred.

Sessions: grouped by the analytics session ID and visitor ID; an anonymous visitor ID is used when the session ID is absent. Only sessions with a pageview enter session-rate denominators.

Active time: the maximum reported cumulative foreground time per session, summed once. The old dashboard looked for active_seconds on milestone events, which actually store seconds. We use those milestones and page-leave active_seconds. This is a minimum: gaps between milestones, lost page-leave events, multiple page loads and visits crossing the period boundary limit accuracy. Visible does not necessarily mean attentive.

Media: deliberate starts exclude autoplay. Milestones use actual played seconds where the player supplies them. Percentages compare sessions that started an item and reached the milestone within the same period; they do not measure individual-play completion or guarantee event order.

Comparisons: adjacent rolling periods of equal duration, sharing the same cutoff. A zero baseline is shown as “New” instead of an invented percentage. Daily visitor counts should not be added to infer unique visitors over the whole period.

Coverage: analytics only includes received Studio-tagged events. Opt-outs, blocked scripts and missed events reduce coverage; bots may still appear. No server-side internal visitor ID exclusions are configured. Browsers opted out through Studio privacy controls do not send events.