Member blog – Teads – Measurement: In an age of viewability how can we standardise technical specifications? The rise of VAST 4.0
by Ryan Cook, Global Director, Programmatic Operations and Ibrahim Ennafaa, Product Manager, Formats Team
What does viewability mean to you? “I would like my ad to be viewable” is far too vague of a term these days. Whilst incredibly important, it’s almost humorous when we think that just 2-years ago viewability was hardly an appetizer at a buzzword dinner table. The importance of viewability is only increasing as digital video increases 85% year on year and costs up to 10x more than display ads.
By definition: “A served ad impression can be classified as a viewable impression if the ad was contained in the viewable space of the browser window, on an in-focus browser tab, based on pre-established criteria such as the percent of ad pixels within the viewable space and the length of time the ad is in the viewable space of the browser.”
The key component here is “pre-established criteria”, which differs for various agencies or individual advertisers. As such, many measurement companies have skin in the game. Each new vendor is offering new ways to monitor viewability in various mediums: mobile applications, mobile websites, desktop, and wearables.
One of the major focal points for the IAB Tech Lab to support the industry is to have a consistent set of standards and structured ability to verify the quality of delivery. As such, there is a prodigious update to Video Ad Serving Template (VAST) documentation for VAST v4.0.
To date, most third party measurement companies are taking advantage of VPAID capabilities to run their measurement code as a verification service. However they also have to deal with the VAST parsing just like a regular ad video player as they are wrapping the client’s original VAST inside of their VPAID.
This is becoming an alarming issue. This utilisation of the VPAID capabilities for this purpose (which is not an IAB specification), is causing delays in ad rendering, increasing latency, ruining user experience and deteriorating campaign performance. In the outstream video space, an increasing number of clients are seeking to measure viewability, which is a good thing, but viewability verification is often activated at each layer of the funnel (primary ad servers, secondary ad servers and DSPs) resulting in having multiple levels of VPAID wrapping. The problem here is that each VPAID mediafile will parse the next VAST instead of simply loading and playing the video as the IAB Tech Lab specifies.
A solution exists and the IAB Tech Lab is already trying to tackle the issue in the VAST V4.0 specs. The code used for ad measurement should stand alone and be part of a distinct VAST element called AdVerification instead of being included in the VPAID mediafile. This alternative allows every supply side platform to implement the third party code properly and guarantee a good ad experience for the end user. Measurement companies will benefit from this as they will be able to focus on data collection without having to implement any video player ad serving feature (i.e. VAST unwrapping, mediafile selection, events tracking etc…). Recent testsbetween Integral Ad Science and Teads showed that a direct integration reduced the discrepancies drastically and provided a clean ad experience (see results here: https://teadstv.app.box.com/s/e9dkfgg397t2yx227p7rcs8wc6nz4a7r) .
The challenge will be to quickly bring together all the stakeholders to finalise the AdVerification specifications and make sure it’s properly rolled out across the ad industry.
The update to VAST 4.0 has further reaching qualities such as offering additional error codes which allow for more granular reporting and troubleshooting on delivery issues.
Bonus: What is a VAST compared to VPAID and classical video mediafiles?
VAST stands for Video Ad Serving Template, it defines an interface of communication between ad servers and video players. It provides all the information necessary for a player to run a video ad. We can have multiple layers of VASTs, called wrappers. Each layer represents an ad server in the funnel.
Inside the last wrapper, we find the mediafile, which represents the actual media element that should be loaded in the player. This mediafile is either a VPAID or a classical video file (mp4, ogg, webm etc …). Now, what is the difference between a VPAID and a classical video file and why VPAID leads more often to issues?
Let’s say the video player is a TV screen which integrates a Bluray player. This TV is owned and controlled by the supply side. The classical video file being the actual Bluray Disc, you simply load it into the TV and press play. If the TV screen is ok, everything should work properly. If not, it’s the responsibility of the supply side anyway and they have everything to troubleshoot. On the other hand, the VPAID is more than a Bluray, it is actually an entire Bluray player with a Bluray disc loaded inside. You have to plug it to your screen, press play and hope everything will work out. The truth is, since you have no access to the inside of the Bluray player, the only thing you can do is press buttons on the remote control and rely on what the player’s screen messages. VPAID stands for Video Player Ad-Serving Interface Definition and brings more freedom to creative vendors allowing them to add interactivity, tweak the video rendering and add rich media elements, but as explained it could also bring more issues. Therefore, a strict implementation of the VPAID specs is a must for a clean ad industry.