About

The Top videos and User engagement reports provide data about content performance and viewer engagement.

Kaltura's current viewership metrics use 10-second interval tracking, which records viewing activity every 10 seconds during playback and provides more precise viewing data.

Downloadable reports can also include metrics calculated using quartile-based tracking, which measures viewing based on milestones reached during playback.

This article defines the metrics available in the Top videos and User engagement reports and explains how viewership is measured.

Access the reports

You can export the Top videos and User engagement reports from the analytics dashboard. When the export is ready, you'll receive an email with links to download the selected reports. The download links are available for 7 days.

Understand viewership metrics

Downloadable reports can include both interval-based and quartile-based viewership metrics.

Interval-based metrics

Interval-based metrics use viewing events recorded every 10 seconds during playback. This is the current method used for viewership metrics in analytics dashboards.

The main interval-based metrics are:

• sum_view_period – Total viewing time.
• avg_view_period_time – Average viewing time per play.
• avg_completion_rate – Average percentage of a video watched by viewers.

Interval-based tracking applies to VOD and live streams, except meetings. Availability of historical interval-based data depends on when the player used to view the content was updated to support this tracking method. For some content, interval-based data can be available as far back as 2018.

Quartile-based metrics

Quartile-based metrics calculate viewing activity based on milestones reached during playback. These milestones represent 25%, 50%, 75%, and 100% of the video.

Quartile-based metrics remain available in downloadable reports alongside interval-based metrics.

Top videos

The Top videos report includes the following metrics for content performance and viewer engagement.

System name	Display name	Definition
object_id	Object ID	Entry ID.
entry_name	Entry Title	Name of the entry.
creator_name	Entry Creator Name	Name of the entry creator.
created_at	Creation Time	Time the entry was created, in epoch format.
status	Entry Status	Entry status, represented by an enumeration value.
media_type	Media Type	Media type, represented by an enumeration value. See Media type values below.
duration_msecs	Entry Duration (Milliseconds)	Duration of the entry in milliseconds.
entry_source	Entry Source	Source from which the entry originated or was uploaded.
count_plays	Total Plays Count	Total number of plays.
sum_time_viewed	Total Minutes Viewed (Quartile Based)	Total viewing time calculated using the last completed quartile. Calculation: Entry duration × percentage of the last completed quartile. Example: A video is 60 minutes long and a viewer watches 50 minutes. The last completed quartile is 75%, so the calculated viewing time is 45 minutes (60 × 75%).
avg_time_viewed	Average Minutes Viewed (Quartile Based)	Average quartile-based viewing time per play. Calculation: Total Minutes Viewed ÷ Total Plays Count. Example: A 60-minute video has three plays. Viewer 1 reaches no quartile, Viewer 2 reaches 25%, and Viewer 3 reaches 50%. The average quartile reached is 25%, so the average viewing time is 15 minutes.
count_loads	Player Impressions Count	Number of times the player loaded on a page. Each player load counts as one impression.
load_play_ratio	Play-to-Impression Ratio	Ratio of plays to player impressions. Calculation: Total Plays Count ÷ Player Impressions Count.
avg_view_drop_off	Average Drop-off Rate (Quartile Based)	Average last quartile reached across all plays. Calculation: Sum of quartiles reached across all plays ÷ Total Plays Count. Example: Three plays reach 100%, 75%, and 50%. The average is 75%, so viewers reached an average of three out of four quartiles.
unique_known_users	Unique Users Count	Number of unique identified users who generated an event included in the report.
engagement_ranking	Engagement Ranking (1-10)	Engagement score ranging from 1 to 10.
avg_completion_rate	Average Completion Rate (Interval-Based)	Average percentage of a video watched by viewers, calculated using 10-second interval tracking.
unique_viewers	Unique Viewers Count	Number of unique users who generated a play event.
sum_view_period	Total Minutes Viewed (Interval-Based)	Total viewing time calculated from events recorded every 10 seconds during VOD and live streams. Meetings aren't included.
avg_view_period_time	Average Minutes Viewed (Interval-Based)	Average viewing time per play for VOD and live streams, calculated using 10-second viewing intervals. Meetings aren't included. Calculation: Total Minutes Viewed ÷ Total Plays Count. Example: Viewer 1 watches 1 minute, Viewer 2 watches 2 minutes, and Viewer 3 watches 1 minute. The total viewing time is 4 minutes, so the average viewing time is 1 minute and 20 seconds.
total_completion_rate	Total Completion Rate (Interval-Based)	Total completion rate calculated using interval-based viewing data.

Media type values

The media_type field uses the following enumeration values.

Value	Media type
1	Video
2	Image
3	Text
4	HTML
5	Audio
6	Show
10	XML
11	Document
12	SWF
13	PDF
201	Live stream
202	Live stream Windows Media
203	Live stream RealMedia
204	Live stream QuickTime

User engagement

The User engagement report includes the following metrics for individual viewer activity and engagement.

System name	Display name	Definition
id	ID	User ID.
full_name	Full Name	User's first and last name.
email	Email Address	User's email address.
unique_videos	Unique Videos Viewed Count	Number of unique videos viewed by the user.
count_plays	Total Plays Count	Total number of plays.
sum_time_viewed	Total Minutes Viewed (Quartile Based)	Total viewing time calculated using the last completed quartile. Calculation: Video duration × percentage of the last completed quartile. Example: A video is 60 minutes long and the viewer watches 50 minutes. The last completed quartile is 75%, so the calculated viewing time is 45 minutes (60 × 75%).
avg_time_viewed	Average Minutes Viewed (Quartile Based)	Average quartile-based viewing time per play. Calculation: Total Minutes Viewed ÷ Total Plays Count. Example: A 60-minute video has three plays that reach 0%, 25%, and 50%. The average quartile reached is 25%, so the average viewing time is 15 minutes.
avg_view_drop_off	Average Drop-off Rate (Quartile Based)	Average last quartile reached across all plays. Calculation: Sum of quartiles reached across all plays ÷ Total Plays Count. Example: Three plays reach 100%, 75%, and 50%. The average is 75%, so viewers reached an average of three out of four quartiles.
count_loads	Player Impressions Count	Number of times the player loaded on a page. Each player load counts as one impression.
load_play_ratio	Play-to-Impression Ratio	Ratio of plays to player impressions. Calculation: Total Plays Count ÷ Player Impressions Count.
avg_completion_rate	Average Completion Rate (Interval-Based)	Average percentage of a video watched by viewers, calculated using 10-second interval tracking.
count_viral	Entry Shares Count	Number of times an entry was shared from the player.
total_completion_rate	Total Completion Rate (Interval-Based)	Number of percentiles viewed per user.
sum_view_period	Total Minutes Viewed (Interval-Based)	Total viewing time calculated from events recorded every 10 seconds during VOD and live streams. Meetings aren't included.
avg_view_period_time	Average Minutes Viewed (Interval-Based)	Average viewing time per play for VOD and live streams, calculated using 10-second viewing intervals. Meetings aren't included. Calculation: Total Minutes Viewed ÷ Total Plays Count. Example: Viewer 1 watches 1 minute, Viewer 2 watches 2 minutes, and Viewer 3 watches 1 minute. The total viewing time is 4 minutes, so the average viewing time is 1 minute and 20 seconds.