Kevin Lenane is CEO of Veenome, a viewability and verification company that works with a variety of clients, including ad networks/exchanges such as Collective and Brightroll. Here Kevin explains how many viewability and verification tools offer an incomplete view of the user experience.
With the seemingly continuous stream of stories around fraud in display and video advertising, the spotlight has also turned to the efficacy of the tools used to catch fraudsters using bots and other automated techniques. The aim of the many measurement techniques – and the main effort of the IAB – has been to define and measure the viewability of the pre-roll ad over time.
That is to say, if we can simply measure the viewers actual screen view on the video over time and calculate this as a percentage, then we should be able to calculate and ensure viewability. This does seem to solve a lot of the problems we’ve been seeing recently with videos 2000-15000px below the fold – often on auto-play and on mute – that get played every time a page opens. Make no mistake these ads almost never get seen. Even at 8000 or 9000px below the fold, it’s a challenge to scroll down fast enough to see the 15 second ad running in the basement.
So this kind of on-screen percentage measurement – where we calculate the amount of viewing time – works well and brings transparency to these impressions by breaking the ad into quartiles and counting whether the user had the majority of the ad in view during each quarter of the length of the pre-roll ad. It’s an effective and simple way to monitor the time element of a video ad while tracking if an ad is really viewable on the screen.
But video viewability, and particularly video verification, require even more measurement in order to give a truly complete picture. Three attributes stand out, auto-play, the presence of multiple videos, and the muting of audio:
Even if a video is above the fold and in view, the use of auto-play changes the intention of the viewer and their interest level in the actual content. This intention is a real and measurable difference in user attention and should be accounted for – regardless of where the video is in-view or not.
We can measure multiple videos being played at the same time on a page using the quartiles system and they could theoretically both register as fully viewed and verified impressions in spite of the fact that the attention level given to either video impression is likely to have been extremely poor. More often we see partial examples of this where we see multiple things playing on the screen at near the same time. The distraction caused by an additional video playing near another is incredibly high and this ought to be accounted for and at the very least measured for.
The activation of audio is another measure of the full “extension” of the ad impression. For the most part, sight and sound will be more powerful than just a muted video.
By only solving one piece of the viewability puzzle, we leave the door open for the bad actors to simply move to the other places where they can gain an advantage. For instance, using the quartile method alone, running 4 muted videos on auto-play right at the bottom of the viewable screen would register 4 valid viewable impressions while still leaving the user a decent experience if they are focused on the content higher up on the page. Incorporating even some of the above measurements would allow publishers and ad networks to evaluate and avoid these sorts of scenarios. In other words, the more complete the viewability and verification standard and toolset is, the stronger it will be.