Putting programmatic video back on the radar
Columnist Alex Bornyakov believes that to receive maximum benefit from programmatic video and monetize every single ad call with the most effective ad opportunity, publishers need to find an intelligent solution to automating the waterfall.
With video expected to feature in 80 percent of consumer internet traffic by 2019, it’s no wonder traditional advertising budgets are being re-prioritized to reflect this. And as the digital video advertising market continues to expand — with double-digit growth expected annually until 2020 — programmatic video has immense potential.
The benefit of programmatic video over traditional video ad buying is how it leverages real-time data to get the right video ad in front of the right consumer, at the right time and in the right place. But despite the bright prognosis, there are plenty of challenges to be conquered before it becomes the publishers’ default trading method.
Many publishers remain wary of a programmatic approach to selling video inventory due to challenges around delivery and efficiency, and they are reluctant to adopt the technology. So what concerns are preventing publishers from fully embracing programmatic video, and what can be done to put it back on the radar?
Barriers to programmatic video
Concerns revolve largely around latency and fill rates. When a user clicks on a video, resulting in an ad call, it can take a number of seconds for the supply-side platform to check multiple demand sources until it finds one that meets the publisher’s floor price, allowing the ad to be served. This delay has a detrimental impact on the user experience and may lead them to abandon their video view.
To avoid too much disruption to the user experience, most platforms set a time limit of 15 to 20 seconds to serve an ad. If a suitable demand source can’t be found within this time period, the opportunity is wasted, resulting in low fill rates. Publishers only receive around 40 percent of CPMs from programmatic video — the rest is spent on value-add services and middleman fees — so they need to monetize every opportunity and maximize their fill rates.
The root cause of latency and wastage in programmatic video is the manual nature of the programmatic waterfall. Although the actual sales process is automated, the software only reacts to human input.
Traditional waterfalls are generated and managed by ad ops teams, and common logic dictates that guaranteed campaigns with high CPMs are placed at the top, and demand sources with lower CPMs are placed farther down. When an ad call occurs, the high-value campaigns at the top are checked first, before the demand sources with lower-value tags.
But by relying entirely on price, this model does not take into account historical latency and fill rates or incidences of time-wasting VPAID (Video Player Ad-Serving Interface Definition) errors, where an empty VPAID is uploaded by the selected demand source and the opportunity is spoiled. A particular demand source may pay high CPMs, but if it has a history of VPAID errors or low fill rates, publishers will not want to give it high priority. The manual waterfall is labor-intensive and leaves the process open to human error — for instance, placing an ad with a broken Flash container or an incompatible player into the waterfall.
What’s the solution?
For publishers to receive maximum benefit from programmatic video and monetize every single ad call with the most effective ad opportunity, they need to find an intelligent solution to automating the waterfall.
Using sophisticated algorithms, real-time programmatic data and historical data, a dynamic automated waterfall can take into account multiple factors, such as previous fill rates, time to impression and likelihood of VPAID errors — as well as CPMs — to decrease latency, increase fill rates and maximize revenues.
An automated waterfall has numerous other benefits for publishers, too. It reduces reliance on manual labor, which is particularly beneficial for rapidly growing businesses and frees up ad ops to focus on value-add areas such as prescriptive rather than retrospective analysis. It also allows waterfalls to be created and regularly optimized for each individual domain source, significantly increasing the efficiency of the process.
Programmatic video provides a huge opportunity for advertisers to target consumers with relevant content at the opportune moment, so the demand for the inventory is strong and ever-increasing. By replacing the classic human waterfall with an intelligent automated solution, publishers can confidently adopt programmatic, decreasing latency, avoiding spoiled impressions and getting the maximum value from their video inventory.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.