It’s widely reported that anywhere from 20% to 30% of programmatic video spend is lost due to supply-side misrepresentation, invalid traffic and low-quality inventory. That’s billions in wasted ad spend – not to mention billions in ad revenue siphoned away from premium publishers – further eroding high-quality video inventory.
If programmatic video is going to mature beyond the “scale at any cost” phase, we need to rethink how we activate and optimize toward higher-quality inventory. Self-declared inventory (accurate or not) doesn’t fully capture multidimensional video inventory, and static metadata snapshots don’t verify both the ad and content aspects of the video experience.
We need to start looking at the deterministic signals that tell us how video ads and the surrounding content behaves on the page to gain a much more relevant, fine-grained understanding of the actual viewing experience.
How the video experience gap grew so wide
It’s easy to blame opportunistic publishers masking low-quality inventory for our current status quo, but we arrived here in a much more organic way.
Over the past five years, programmatic fundamentally democratized access to video. But to scale quickly, the industry developed labels as a proxy for ad experience. “Instream,” for example, implied a pre-, mid- or post-roll ad within real editorial video content. “CTV” implied a premium, lean-back viewing experience on a television screen. And premium OLV environments were expected to mirror that same immersive, video-first experience within web and mobile environments.
But as programmatic video supply has expanded and supply-side competition has intensified, that connection between the label and the lived ad experience is now far more ambiguous.
Inventory categorized as instream isn’t always anchored to meaningful editorial video content. In fact, instream-labeled inventory often sits on sites that don’t even have a valid video player to serve the…