User analytics plugins

This section describes Kusto extensions (plugins) for user analytics scenarios.

Scenario	Plugin	Details	User Experience
Counting new users over time	activity_counts_metrics	Returns counts/dcounts/new counts for each time window. Each time window is compared to all previous time windows	Kusto.Explorer: Report Gallery
Period-over-period: retention/churn rate and new users	activity_metrics	Returns dcount, retention/churn rate for each time window. Each time window is compared to previous time window	Kusto.Explorer: Report Gallery
Users count and dcount over sliding window	sliding_window_counts	For each time window, returns count and dcount over a lookback period, in a sliding window manner
New-users cohort: retention/churn rate and new users	new_activity_metrics	Compares between cohorts of new users (all users that were first seen in time window). Each cohort is compared to all prior cohorts. Comparison takes into account all previous time windows	Kusto.Explorer: Report Gallery
Active Users: distinct counts	active_users_count	Returns distinct users for each time window. A user is only considered if it appears in at least X distinct periods in a specified lookback period.
User Engagement: DAU/WAU/MAU	activity_engagement	Compares between an inner time window (for example, daily) and an outer (for example, weekly) for computing engagement (for example, DAU/WAU)	Kusto.Explorer: Report Gallery
Sessions: count active sessions	session_count	Counts sessions, where a session is defined by a time period - a user record is considered a new session, if it hasn't been seen in the lookback period from current record
Funnels: previous and next state sequence analysis	funnel_sequence	Counts distinct users who have taken a sequence of events, and the previous or next events that led or were followed by the sequence. Useful for constructing sankey diagrams
Funnels: sequence completion analysis	funnel_sequence_completion	Computes the distinct count of users that have completed a specified sequence in each time window