Programmatic Video Advertising, Explained Plainly
The video ad supply chain in plain terms — who the parties are, where the money goes, and which levers a publisher genuinely controls.
Programmatic video is opaque, and much of the available explanation comes from parties with an interest in it staying that way. Here is the plain version.
The chain
- You have a video slot and a visitor.
- An ad server decides what fills it — a direct-sold campaign, or an auction.
- An SSP takes that auction to buyers.
- DSPs bid on behalf of advertisers.
- The winning creative is returned and played.
All of it happens in a few hundred milliseconds, and every hop takes a percentage.
Where the money goes
The gap between what an advertiser pays and what a publisher receives is substantial once every intermediary has taken its cut. Chasing that gap is the industry's favourite activity and rarely a publisher's best use of attention, because the terms are set by parties with more leverage.
Terms worth knowing
- Fill rate — the share of requests that returned a paying ad
- Instream vs outstream — an ad inside video content, versus a player injected into an article
- Header bidding — soliciting bids in parallel before calling the ad server, to widen competition
- Floors — the minimum you will accept; set too high and fill collapses
- ads.txt — a public file naming who is authorised to sell your inventory
What you actually control
Not rates, mostly. You control inventory quality and volume: how many video impressions you generate, whether placements are viewable, and whether the surrounding experience keeps people willing to load your pages.
Which is why the impression side of the equation deserves more attention than it gets, and why a slow player is a revenue problem rather than merely a technical one.
On ad load
The honest position: every increment raises revenue today and reduces the audience willing to return. For streaming specifically, the formats inherited from broadcast are the wrong starting point.