• by Mendel Senf , Op-Ed Contributor, May 5, 2014

There have been a lot of chatter and criticism directed at the often extreme measures that are taken to measure online advertising. Reporting ROI for clients based on clicks and engagement rates is absolutely expected on the client side but surely, critics moan, digital is taking measurement to the extreme? It could be said that up until now, outdoor advertising has been an example of a completely opposite, less measured format – essentially we can't accurately tell who sees a billboard or how many people took a sideways glance at a poster and clocked the company logo, ready to Google later.

So why do we need to find out exactly where eyeballs are looking on the web? Brand marketers today are trending towards buying an audience online in an automated fashion rather than going to buy online ad space directly from a media owner like TheGuardian.com or Gardeners World, as they would have done with a paper order before. This practice of tracking down users and buying impressions in a manner akin to buying stocks and shares is becoming the norm, thanks to the effectiveness of ad exchanges to bundle audience pools together for sale. However, the sheer scale of these transactions — at the most extreme you might be targeting 20,000 users on 20,000 different websites — means that going, looking and doing an audit of where your ad appeared is impossible in many cases.

Enter viewability. We're now looking at a digital media buying landscape that's wholly altered from the world we entered a decade or so ago. This week, the Media Ratings Council (MRC) in the U.S rethought its advisory against selling digital ad space using viewability metrics, which is great news for businesses trading in “ethical” media — both on the publisher and advertiser sides, but also for companies that have developed technology to weed out the bad-egg sites driving “fake” traffic and non-viewable ads, such as Integral Ad Science, Nielsen and Meetrics.

A lot more housekeeping will need to be done in the coming few months, and, of course, creating a standard measurement for viewability based on, for example, whether an ad is 100% in view for 50% of the time will mean different outcomes for the various ad formats available currently. The MRC hasn't given its blessing to a video metric as yet, but for mobile video inventory we could assume that a mobile video would be 100% viewable due to the fact that it takes over the user’s screen, so rather than identifying poorly placed ads which appear below the fold of a web page, for example (as with display and desktop video), viewability metrics could prove to be a powerful tool at weeding out "fake click" mobile video sites.

To truly understand and utilize viewability measures, they will need to be analysed in conjunction with other metrics and data. For instance, we should look at metrics that are derivatives of viewability, like “in view for more than five seconds” and “the average time that an ad was in view”. These metrics help us more accurately measure “actual viewability”. It also important to interpret the data collected from your user. Who they are, their past behavioural history and where and when they viewed the ad all play pivotal roles in helping us understand what an appropriate viewability metric is. 

Viewability as well as bounce rate could prove a killer combination when measuring the reach and effectiveness of native ads that are moving towards being bought in real-time. There is a huge amount of potential but one thing is for sure: advertisers want viewable ads, and this demand will create a new meritocratic economy. Mark my words.