How dynamic ad insertion works
The video carries markers for its ad breaks but no ads. When a viewer reaches a marker, a request goes to an ad server describing the viewer and the content. The ad server returns the ads to play, and they are inserted either in the player (client-side) or in the stream (server-side).
Why it earns more than fixed ads
- Targeting: advertisers pay more to reach a defined audience
- Freshness: a title watched next year carries next year's campaigns, not expired ones
- Frequency control: one viewer is not shown the same ad repeatedly
- Fill: unsold breaks can be passed to programmatic demand or skipped
- Measurement: every impression is counted and reported
DAI, SSAI and CSAI
Dynamic ad insertion is the principle; SSAI and CSAI are the two ways of carrying it out. In some markets, DAI is used loosely to mean the server-side method. The opposite of all three is a burned-in ad, which is part of the video file and identical for every viewer.
What DAI needs
- Ad break markers in on-demand titles and live streams
- An ad server or ad network to make the decision
- Standard tags, normally VAST and VMAP, to describe ads and breaks
- A player or stitching service that supports insertion on each device
- Consent handling for the viewer data used in targeting
Last reviewed