The Open Web's Video Quality Problem

Most of the video impressions being bought today aren’t delivering what buyers think they're paying for. Not because buyers are careless, but because the industry has leaned heavily on a classification system that was designed to simplify buying, not measure real viewing experience. 

For years, Instream and Outstream labels have acted as shorthand for quality. Instream, in particular, has become the default choice. It feels safer and more premium. But when you look at what actually happens on the page, that confidence fades. 

DSP labels are broken

The IAB’s Instream definition focuses on how a video player appears at page load: primary content, sound on by default. It sounds reassuring, but it says little about what happens once a user starts scrolling. A video can qualify as Instream but still collapse into a sticky unit, play muted or continue playing off-screen. Until recently, buyers haven’t had the visibility to detect this, leading to wasted budget and completion rates that overstate real exposure.

At the same time, Outstream (representing around 90% of supply) is excluded by Instream-only strategies, removing a significant pool of genuinely high-quality environments. When most budgets compete for less than 10% of inventory, demand concentrates, auctions intensify and CPMs rise. 

Those higher prices are often interpreted as quality, when they’re simply a function of scarcity.

DSP controls are built around declared attributes like player size at load and format taxonomy. Useful, but static. They fail to reflect how video behaves throughout the session and when the same platform both defines quality and optimises delivery… then buyers are left without an independent view of what’s really happening. 

‍

15% of videos continue playing after users scroll

At Picnic, PIQ was built to simulate real user experiences across the open web to enable advertisers to make data-backed inventory decisions. When we applied that same lens to video inventory using our Quality Video product, we found some concerning patterns:

• 30% of video players collapse into smaller units during the user’s session.
• 15% of videos continue playing after the user has scrolled them entirely out of view.
• 12% of sites run multiple videos simultaneously on the same page.

Completion rates and format labels simply don’t capture this yet these behaviours directly determine whether an impression had a genuine opportunity to be seen and heard.

‍

“Our analysis shows that video impressions don’t always behave the way buyers assume. A player can technically meet a DSP label at load, yet function in a way that changes the nature of the impression entirely.

That matters because experience is what determines value. If video is going to command premium investment, it needs to be measured and priced based on how it actually behaves on the open web.”

‍

Nick Evans, Product Manager at Picnic

‍

Quality Video: Buy video based on how it actually behaves

Quality Video is Picnic's answer to the industry’s video quality problem. It's a first-of-its-kind video curation solution that evaluates inventory based on real player behaviour.

Built using PIQ's Inventory Intelligence, Quality Video scores video supply using a range of signals.

These signals reflect how players behave throughout the user journey such as: average player size over time, pause-out-of-view logic, integration behaviour (fixed, sticky, collapsing), or whether video ads play concurrently on the page. 

Rather than starting with “Is this Instream?’, Quality Video asks “Does this create a genuine viewing opportunity?” That shift opens up access to high-quality Outstream inventory that has historically been excluded while also challenging the assumption that all Instream is inherently premium.
‍

How Quality Differs Across Inventory
‍

PIQ measures real on-page signals, exposing clear differences between low, standard and premium quality inventory.

‍

What Quality Video Gives You That Labels Can't

More meaningful exposure

Spend is focused on impressions with a genuine opportunity to be seen, heard and engaged with.

Reduced waste from low-quality supply

Poor-quality inventory is able to be filtered out before spend begins, reducing wasted spend.

Transparent, confident video buying

Clear visibility into how video actually behaves based on real player signals, giving buyers control over what “quality” means for their campaign.

Better alignment to brand objectives

Video appears in environments that support genuine consideration and brand impact — not just environments that look good on a post-campaign report.

The Bottom Line

Video’s quality problem won’t go away on its own. The labels are blunt and the metrics are misleading. Quality Video changes that, grounding buying decisions in how ads actually behave on the page instead of how they're labelled.

If the industry wants video investment to drive real outcomes, it has to measure what actually happens on the page.

Want a free health check of your current video site list? Get in touch here.

‍