"On TV And Video" is a column exploring opportunities and challenges in advanced TV and video.
Today’s column is written by Michael Tuminello, vice president of solutions and partnerships at Innovid.
Believe it or not, video ad standards are almost 10 years old. VAST 2.0 and VPAID 1.0 [PDF] were drafted in 2009 and, generally speaking, are still in use today. Incremental updates have mainly addressed publisher concerns and won only partial adoption.
These standards predate the rise of programmatic advertising and digital advertising’s migration from the desktop to mobile and over-the-top (OTT) devices. This means the video ad ecosystem is running on standards designed for a minority use case: direct-bought desktop video player impressions.
VAST 4.0, introduced about two years ago, was meant to better accommodate OTT and mobile, but because it never addressed loose ends related to interactivity and verification, publishers largely didn’t adopt it. VAST 4.1, introduced last week, aims to finish what VAST 4.0 started.
The core change is to separate the video asset from any code and ultimately split what is now VPAID into two separate APIs – one for measurement (Open Measurement Interface Definition) and another for interactivity (as yet unnamed, but temporarily referred to by the IAB as VPAID-I).
We’ve seen a shift in the digital ad business. When VAST 2.0 and VPAID 1.0 were drafted we were solving a quantity problem: How do we scale the delivery and measurement of video ads to meet demand?
Now, having solved (or perhaps oversolved) the quantity problem, we are facing a quality problem: How do we give advertisers confidence that the value they get from digital ads is not offset by fraud or brand-safety issues?
In brief, we need to continue to enable quantity in a multidevice world, while ideally doing what we can to address quality.
Quantity: More scale for currently available functionality
VAST 4.1 provides for better delivery to mobile and OTT and improved viewability measurement.
Better delivery to mobile and OTT: Rather than being wrapped in code, the video asset will be directly accessible.
Mobile video companies retrieve the video asset in advance to eliminate playback delays, and broadcasters stitch the video into a single stream (server-side ad insertion) for similar reasons. VAST 4.1 closes all gaps required to support both scenarios.
Viewability measurement in new inventory: Up to this point, viewability measurement required VPAID, a code layer intended to support interactivity and later co-opted for measurement. Now VAST 4.1 is compatible with the Open Measurement Interface Definition and the OM SDK, so VAST inventory, including in-app, will be able to be measured without VPAID on supporting publishers. The IAB has also outlined how the Open Measurement Interface Definition and OM SDK can be supported in earlier versions of VAST.
Expanded companion use: The use of VPAID may shrink once viewability can be measured using VAST 4.1. The new standard blesses the use of the VAST companion ads to bring interactivity and dynamic functionality to VAST inventory without VPAID. The companion asset has been part of VAST since VAST 2.0 and was originally used as a banner ad alongside a desktop media player.
Mobile pubs intelligently repurposed the same asset as a full-screen end card that would allow consumers to interact with a brand after watching a video ad. Often used to drive app downloads, the format has so far flown under the radar of brand advertisers. Companion end cards are widely supported in mobile in-app inventory already; VAST 4.1 supports the end card and future creative uses of the companion.
Quality: New initiatives to improve transparency
The original VAST standard contained little to address quality issues, which have begun to be tackled by other standards and initiatives such as Ads.txt. Changes to the communication between the ad server and publisher can help to improve quality.
A few items have been introduced to VAST 4.1 that can help if adopted.
Standardized macro support: Macros are used to pass data from the publisher showing the ad to the ad server delivering the ad. The industry has widespread support for macros, but they are unstandardized, meaning macros differ from publisher to publisher. VAST 4.1 suggests a broad set of useful data points and standard macro names to request them across any publisher.
Ad-serving ID: VAST 4.1 proposes the creation of a unique ID every time an ad is served to be included on outgoing pixels to improve discrepancy troubleshooting and compare data from different systems.
How will VAST 4.1 affect me?
The first thing for everyone to note is that standards adoption gradually. There is no need to panic as changes will be phased in over time. Look to your partners to stay informed and push for changes that benefit you.
Publishers can benefit by making more VAST inventory measurable to attract buyers. They can gain premium units in VAST inventory by supporting the VAST end card and other new VAST creative options. Publishers moving to server-side ad insertion now have a standard to work with. Supporting any standards update is an investment for pubs, so buyer demand ultimately determines support.
Buyers win by now having a way to get measurement in VAST inventory, starting with in-app inventory, which has been largely unmeasureable, without requiring full VAST 4.1 support from publishers. Buyers running a lot of mobile video now have the option of driving additional performance and engagement in VAST inventory for the first time, using VAST-based creative formats, starting with the end card. Buyers who insist on VAST 4.1 support from publishers can also start to get a lot more data straight from the player and gain better ability to track individual impressions across multiple systems without PII.
In general, it is time for buyers to “lean in” to ad standards, to pick up the quantity baton from the sell side and determine what additional changes can be made to support quality.
The industry as a whole should also consider whether it wants to embrace VAST 4.1 and stop using older versions, even if it’s easier to just keep using VAST 2.0, which still technically works in a good chunk of available inventory, despite showing its age. A single version means we can avoid day-to-day trafficking confusion and move forward with a baseline better adopted to our current multidevice ecosystem.