Exits Before Video Start

Exits Before Video Start (EBVS) counts the attempts that ended before video start without a reported fatal error during the selected interval.

Updated 2026-07-27 exits-before-video-start

Note: The Wait Time Before Exit metric measures the time viewers wait before exiting the video during EBVS.

This metric shows the sum of Attempts  that ended before video start without a reported fatal error, within the selected interval. 

If a fatal error is reported when the video terminates before starting, the video termination is counted as a  Video Start Failures (VSF) instead of EBVS.

With Client-Side Ad Insertion (CSAI), pre-roll ads are included in the video startup and exits during these ads contribute to EBVS.

With Server-Side Ad Insertion (SSAI), pre-roll ads are not included in the video startup and exits during these ads do not contribute to EBVS.

For more details about CSAI and SSAI sessions, see Metric Events.

As a percentage, this metric shows the percent of viewing Attempts that experienced EBVS

EBVS is an 'Interval' type metric and is counted in the interval in which exit occurs. For more information about Conviva metric intervals, see the Metric Dictionary.

Exits Before Video Start is available in the following formats: 

  • Exits Before Video Start: the total number of EBVS during the selected time period.
  • Exits Before Video Start summary: the total number of EBVS during the selected time period for the top content assets. 
  • Exits Before Video Start time series: a time series for the selected time period, where each point is the number of EBVS during that specific interval.

Note:  If you use an Audio stream or a combination of Video and Audio streams, you can also view this metric as Exit Before Stream Starts (EBSS). To switch the metric measurement between video and audio streams, contact your Conviva representation or Conviva Support.

Abandonment

Abandonment occurs when a viewer exits a video before the video start with a wait time greater than the SPI threshold setting, 10 seconds for SPI Good and 8 seconds for SPI Best. Abandonment appears as an improvement opportunity in the Overview dashboard. Based on the SPI threshold setting, EBVS Wait Time, Abandonment also appears as a metric in Trends and other dashboards. For more details about SPI settings for abandonment and analyzing these improvement opportunities, see Overview dashboard.

Distribution

Wait Time Before Exit distribution:  a distribution of the Attempts that resulted in wait times across a series of predefined buckets. The buckets represent different ranges of Wait Time (seconds), and each bucket displays the number of Attempts with EBVS for the wait time ranges. You can use this distribution to help determine the amount of wait time users tolerated before exiting the video start. 

You can use metric filtering to diagnose only the sessions that had high percentage of dropped frames. For more details on metric filters, see Metric filtering.

Related Details

Metric Details

Description

Dashboards and Metric Presentation
  • Overview
  • Real-Time
  • MetricLens
  • Viewer Sessions
  • Diagnostics
Dimensions Use MetricLens dimensions to filter this metric by metadata segments for enhanced root cause analysis. For example, limiting data to specific ISP and player dimensions can narrow the root cause of higher EBVS to a known provider and player type.
Diagnostics

The Diagnostic page presents detailed information about the Exits Before Video Start in the Analysis window. For more information, see .

Manual Alert Support

Yes. Configure alerts to receive proactive notification when EBVS issues occur. For more information, see Alerts.

AI Alert Support Yes. AI alerts offer proactive anomaly detection and instant alerting to help shorten time to resolution. For more information, see AI Alerts.
Metric Adjustments

None.

Note: Revived sessions that play successfully also contribute to Attempts, Plays, and Ended Plays. Revived sessions do not contribute to VST, EBVS, and VSF. For more details about revived sessions, see Session Timeout/Expiration.

For more metric definitions, see the Metric Dictionary.