Skip to main content

Types of insights

Association insights

Association insights tell you what subgroups of the data are underperforming in your metrics relative to the rest of the population. They also tell you how much the metrics could hypothetically improve by addressing the inefficiency.

Screenshot showing an example of an insight and they types of information shown in each insight.

In the example insight above, a low Order Cancellation Rate is desirable. The Order Cancellation Rate considering all 242K orders in the data set is 16.2%. However, there are 5.54K orders for a particular customer GlobalHorizon Enterprises where there has been a Sales Order Regulatory Hold. The Order Cancellation Rate of this subgroup is 90.0%. This is much higher than the global average, suggesting there is room to improve. The improvement in the Order Cancellation Rate is estimated to be from 16.2% to 14.2%.

Cohort trend insights

This feature is currently available as a Private Preview only

During a Private Preview, only customers who have agreed to our Private Preview usage agreements can access this feature. Additionally, the features documented here are subject to change and / or cancellation, so they may not be available to all users in future.

For more information about our Private Preview releases, including the level of Support offered with them, see: Feature release types.

Cohort trend insights reveal how your metrics change over time, as well as which cohorts drive (change in the same direction) or detract from (change in the opposite direction) those trends.

Screenshot showing an example of a cohort trend insight in Insight Explorer.

There are several important considerations for the using cohort trend analysis:

  • Only metrics utilizing an AVG aggregation type are supported.

  • An event log is required to construct the time series being analyzed.

  • This time series is constructed based on the timestamp of the first event per case.

  • Metrics using FILTER_TO_NULL will have limited support which can lead to issues resulting from the inconsistent propagation of filters.