Online display ads have been evolving quite quickly. From the annoying pop ups, to the more palatable rectangular box on the ride hand side of the screen and to ads blending into your Facebook news feed, there have always been questions regarding what is the best way to judge the effectiveness of these ads. Clicks and Clickthrough rate were the lead indicators for most of the display ads lifespan. Recently viewthroughs have come into play and allowed marketers the ability to judge an ad on not only if someone clicked on it (which is rare) but also to see what users that were exposed to the ad did after “seeing” it. This led to different key performance indicators such as Cost per Click to Cost per Landing and Cost per Response (users who land from the click or who were exposed to the ad and managed to find their own route to the client’s site).  Ad viewability is about to change the way we examine performance of all online banners for the better.

Integral Ad Science allows marketers to assess their media buys on multiple different factors, one being on whether or not the ad that was purchased was actually viewed by the user. They determine whether the ad was viewed by seeing if the ad was in view for one second or more. Being in view means that 50% of the ad or more was on the viewable page for the user. This definition of viewed is quite generous as it is usually unlikely that the client has their logo on both the top and bottom of the ad. Even if the ad was deemed viewable, there is still a good chance that the ad was never actually seen or if it was the user was unable to attribute it to any company.

Now ad viewability is going to help the industry in multiple ways – one it is going to start battling the issue of digital marketing fraud. Two of the many types of fraud would be: clickfarms, where bots are made to click on hundreds of thousands of banners in order to make money, and ad stacking, where sites place multiple ads on top of one another in order to sell more ad impressions and when clicked all the ads are clicked at once, but the landings are not counted. Ad viewability will be able to help with this fraud because first and foremost it will allow the marketer to see if the ad was ever in view, and if it wasn’t it explains why it might never have been for example if fraud stacking occurred. This function in Integral is one of the most important aspects, as it allows marketers to go back to the sites or vendors and explain what is going on. While I have yet to make buys off of ad viewability, this is something that will start being a focal point in buys. The whole industry will soon make buys off of target audience, CPM price, and having a certain percentage level of ad viewability. This will aid in what is often the case of large deliveries near the end of the month where different sites buy traffic to meet certain levels they sold to clients.

Ad viewability allows for more than just whether the ad was seen, it also allows us to see in an anonymous way who saw the ad. Was the user that saw the ad really a person or was it a bot? Was that user actually in the United States or was it someone from a clickfarm in the Philippines or India? We often just take the vendors word for where the traffic is coming from, what this will allow the industry to is to appraise whether the buy was actually worthwhile. If all we have to see is whether the users click on the ad or not, there is not much to differentiate vendors from one another. When we were able to add landings and viewthroughs this allowed us more data to determine which vendors were outperforming one another in terms of cost metrics. Now we can add another layer for whether the ads we paid for were actually seen on the page or even hit the right target. It is somewhat straight forward on how to judge two different buys on whether they drove sales, as we can track if the users that clicked on the ad or were exposed to the ad actually converted. However, with the adviewability we can now see whether those users that were exposed actually had a chance of seeing the ad, or if they would have found their way to the site regardless. This will be determining factors in awareness buys as many companies buy ad space on sites that fit their potential clientele, now that potential clientele might be going to your site regardless, and if the ad viewability of your extremely expensive buy is only 30% but your viewthrough is very high, we might have a case where we are hitting the right clients but spending unnecessary dollars.

This is an exciting time for data analysts like myself as it allows us more data to play with. With more data comes more learning’s, and more points from which to judge different site buys. In the end it will only help get better media plans because we will buy the best ad space to hit our clients on the site where the KPI’s are best met. It will also start to reform the industry and remove the fraud that has plagued it in recent years. Times are changing in the industry and for the better.

Sebastien Damas
Just Media, Inc.