New Statistics used a new data model that we have developed to provide much more accurate on how your visitors experience and interact with your Flipbooks.
However, the New Statistics data model can make your Flipbook statistics and values appear significantly different from how you have experienced them before, specifically metrics such as:
average time spent
average page views
We understand these changes, can be concerning. This guide aims to shed light on how we calculate these metrics in New Statistics, compared with how we did in Legacy Statistics.
Why are these statistics so different in New Statistics compared to Legacy Statistics?
While we never tracked pages viewed or time spent, without consent, in the Legacy Statistics data model, we mistakenly calculated the averages by dividing with the full number of sessions, instead of only the sessions where we had consent. This affected the accuracy of these averages. In the New Statistics data model, we have now corrected the method to ensure more reliable metrics.
Metric: Sessions
The number of sessions metric isn't impacted by consent management. Any small differences in session counts between old and new data are due to improved accuracy in session tracking.
Metric: Avg. time spent
In the data model used for Legacy Statistics, we identified two inaccuracies:
Tracking inaccuracy: In Legacy Statistics, time spent was incorrectly doubled per spread. For example, 8 seconds on pages 2-3 were logged as 16 seconds, inflating session times. These values were then aggregated across spreads viewed in a session, which ultimately resulted in an inaccurate average time spent metric.
Consent Management inaccuracy: For Flipbooks with a consent management solution, we divided total time spent for sessions with consent by total sessions, including non-consenting ones, skewing the metric.
In New Statistics, we now track accurate session lengths for individual pages, eliminating the doubling error. For mobile devices, time spent on individual pages is tracked for graphs. Consent-based averages are calculated only from consenting sessions. These updates greatly improve the accuracy of the average time spent metric.
Metric: Avg. page views
The Average Page Views metric was also affected by the aforementioned Consent Management inaccuracy.
We recorded page views for consenting sessions correctly, but in the Legacy Statistics data model, we mistakenly divided this by the total session count, including non-consenting sessions, lowering the average. To fix this, we now calculate the average using only consenting sessions in the New Statistics data model, ensuring a more accurate reflection of user engagement.
Changes in the Average Page Views metric are due to the correction for the Consent Management inaccuracy in the New Statistics, that were previously reflected in Legacy Statistics. Other factors that can account for changes in this percentage are further optimizations in page view and session tracking.