Ad-viewability is an old advertising concept that still applies to new digital media today.
The Internet, as a media, cumulates the strengths and weaknesses of print and television: just like the press, the Internet allows simultaneous consumption of content and advertising. This allows advertising messages to remain displayed for quite a long time while the user is reading textual or watching video content. So it is not surprising to notice that the market uses the word “impressions” to talk about online banners.
But with this strength, comes the constraint that there is no guarantee that the ad is actually viewed. With print, it is difficult to know what pages a magazine or newspaper was actually opened at. On the web, difficulty comes from the size of screens that do not allow the full page to be totally viewed.
As such, for the first time in the advertising industry, a banner can be only partially viewed on the web. For instance, the bottom of the banner may be out of view on the screen of some users.
Using “impressions” for online ads may be misleading; most online ads are not static images, like in print. They are usually animated images, flash animations, or videos. These formats share an important characteristic: their message has a duration.
So, just like on television, the duration of the message and the duration of the exposure to this message matter.
For every media, in every market, some data is necessary to plan, measure and optimize advertising campaigns. TV uses panels and data from set-top boxes while print relies on circulation, surveys, and other metrics.
This is where the Internet shows its biggest strength: technology makes it the most measurable media.
It is possible to measure if an individual ad-impression is actually in view.
In 2012, the US IAB came to a definition that addresses two dimensions: geometrical and chronological.
An impression is viewable when at least 50% of its area is in-view during at least one second.
This definition sets a minimum threshold that both the buy and sell-side agree on. It does not deal with the concept of ad-efficiency.
This definition looks simple at first glance. So why did the Media Rating Council wait until late March 2014 to allow the market to trade on viewable impressions?
Such a definition is actually too simplistic, so many other questions have emerged:
What is the area of a banner that can expand?
Can the same definition apply when the dimension of the ad is larger than the screen itself?
What if the banner object (image of flash) is not loaded? Is it ok if the placement where the ad is supposed to be is in-view?
What happens when the banner is contained in an iFrame that blocks information from the page?
Are two views of 0.5 seconds equivalent to 1 second?
What if the banner is viewable on the computer’s screen, but nobody is actually viewing the screen? What about auto-refresh pages?
What if the banner is loaded by a robot on a farm of servers?
What if the banner is viewed by someone who is paid to view pages and click on links?
In other words, ad-viewability is a simple concept that solves a big problem, but constantly raises some new questions.
Some questions have already been answered by the IAB:
The one second must be consecutive.
The banner object must be loaded in the page.
Big banners may need to have only 30% of their area in-view.
Robots must be excluded from viewable impressions.
At AppNexus, we benefit from the years of experience of the Alenty team. Alenty was the first company to actually measure ad-viewability, back in 2007.
All these questions have been raised years ago. Instead of waiting for the market to answer them, AppNexus will propose answers and solutions with some ideas in mind:
Ad-viewability must go further than defining a potential exposure. All possible efforts must be made to ensure that a human is actually exposed to the ad.
All new formats and devices must be measured in a consistent way.
The goal of these general rules is to ensure that ad-buyers will get the best quality of inventory for their campaigns, and that publishers with quality inventory will get better prices than publishers with poor quality or fraudulent inventory.
With our experience and our forward-thinking ideas, AppNexus will lead the market to evolve to an era where only viewable impressions matter.