Views and Viewability

06 July 2015

The debate around what constitutes ‘viewability’ has been around for a while, which is strange as in one sense it’s really pretty obvious, even non-contentious.

If an ad cannot be viewed then it has zero value. It therefore stands to reason that advertisers will only pay for ads that can be viewed.

The complications come when this simple concept gets confused with other related notions.

The first such confusion is between viewability and what has actually been viewed. For an ad to be viewable, it has to be there, in its entirety in front of an audience.

Whether the audience then decides to actually engage with the ad is a whole different thing and is of course affected by all sorts of external criteria (such as time of day, context of placement, mindset of the person doing the viewing, as well as the quality, relevance and appropriateness of the creative work).

Viewability as a concept also has nothing to do with the audience delivered. If the ‘audience’ that has allegedly been exposed to the ad turns out not to be human but a bot then of course that’s a serious matter, but it’s a different serious matter.

For an ad to be viewable it has to be visible. If a part of it is not visible, because the viewer doesn’t scroll down that far, or it’s below the fold, or whatever, then it’s not worth the same as an ad that is wholly in view.

The wooliness in this debate is hidden away in that little word ‘wholly’.

The Internet Advertising Bureau has decreed that an ad is considered viewable if 50% of that ad is in view for at least one second.

I cannot think of a single occasion when the creative team behind the work have said: “You know, as long as 50% of this thing is in the consumer’s vision for one second, it’s done its job.” It’s clearly a nonsensical concept, and one that (it has been made clear publicly) is unacceptable to several major agencies and many large advertisers.

It is worth noting that at least the IAB has done something – although their comments at the time that this is simply the first stage of a work in progress don’t seem to have led anywhere.

Many digital sales guys defend the IAB definition by comparing online advertising with more traditional media forms, like TV, or print. But that comparison simply doesn’t stand up.

If viewability is all about whether the ad is there at all to be seen, then every TV ad is viewable. It’s rare (although I concede it does happen once in a while) for a TV ad to be cut short through some technical broadcast malfunction. When that does happen, the advertiser is inevitably immediately refunded.

Same with print, if an ad is in a magazine, it’s there to be viewed. It’s viewable. If it’s not there for some reason, it’s not charged for.

Whether TV or print ads actually are seen is another question, one that has been addressed by many research studies over literally decades. Media planners factor in all sorts of placement criteria to increase the odds of their ad being seen, but the one factor that is constant is that the ad is at least there to be seen, should the audience choose to view it.

It seems to me that it is essential to retain a degree of flexibility when it comes to reporting on viewability; ‘one size fits all’ is not the answer.

For performance advertisers the key metric is response, however measured. An ad being fully in shot for a reasonable period of time (certainly long enough to allow the consumer to take some action – like click) will naturally aid performance but at the end of the day the campaign will be judged on the direct response generated.

Brand building campaigns, of the sort favoured by most large advertisers will ultimately be assessed on metrics like awareness, favourability, satisfaction scores, propensity to buy and ultimately sales. These ads are typically pre-tested before they ever appear, so that any misinterpretations and weak communication points can be ironed out.

These advertisers select a medium for all sorts of reasons – but in every single case it is assumed that at least the ad will be put in front of the selected audience as intended. This is why many big spenders, including Unilever have stated that only 100% viewability is acceptable.

Clearly there is a pay off: the higher the bar, the lower the measured audience but the greater the end value.

As publishers increasingly build their direct dialogues with advertisers there is a greater importance to be attached to understanding what drives business success. Going beyond the media numbers if you like. Ensuring an ad is there to be viewed would seem to be the basic building block onto which more relevant metrics can be stacked.