In the last 6 months there have been a number of high profile news stories which bring the measurement of what we do as digital marketing into question more than ever, particularly in the world of display and programmatic advertising.
Facebook admitted to misreporting numbers and are still bumbling their way through a PR nightmare, one which will leave them no choice but to let third party verification partners into their walled garden. Twitter too has become embroiled in reporting issues which have resulted in refunds to advertisers. And more recently the much refuted ‘Methbot’ discovery by White Ops has once again brought Ad fraud to the forefront of discussions.
Individual reports and issues aside, one measurement topic which is going to come to a head for many advertisers in 2017 is that of Viewability.
Since the IAB announced their standard for Viewability in 2014 it is a topic which is never far from the agenda when discussing display advertising, and especially programmatic display where you trust technology to tell you where your ad appeared.
Advertisers want to know that their ads are actually being seen, after all what impact can they have if they aren’t being viewed by people?
But does Viewability actually matter? Of course the simple answer is yes, especially if you are measuring performance based on post impression conversions. But the level to which it is important is a slightly more complicated matter.
What we are seeing in the market right now is quite a natural, extreme reaction to industry concerns about Viewability. Group M are claiming they will only pay for 100% viewable inventory. Advertisers want to trade only on viewable CPM (VCPM) metrics. And a number of publishers are out in the market pushing 100% viewable formats.
But let’s look at the reality for a second and the challenges around Viewability:
The IAB definition of Viewability is 50% of the ad in view for one continuous second. A pretty low benchmark if we are honest. Half the ad for a 1 second pause is hardly what anybody would consider a level which means the user is viewing or engaging, but it is at least a start. What this definition does is mean you could cut out below the fold and out of sight placements from your campaigns that may be cookie spraying and claiming conversions.
What this definition also means is that Viewability is relatively easy to game for publishers. Ads above the fold will have a strong Viewability score, regardless of size. Any sort of interstitial or skin (assuming it is not on a huge page) will also have a great Viewability score.
As advertisers and agencies start to demand higher Viewability standards then the temptation is there for publishers to force the issue. Forcing people to view an ad before consuming content. Interstitial style ad formats with delayed skip buttons. Hidden ads in ‘viewable’ locations. Lots of ways to drive Viewability scores up, but not in a way which is going to be user friendly and drive a positive opinion from the advertising consumer.
Viewability vs cost
As with most things in life, with greater quality increases so does cost. Those who associated programmatic display with lower costs are now the ones shocked when their Viewability is poor.
As Viewability rates rise, so does the cost, so there is a trade-off to be made. It’s on the agencies and clients to find the tipping point where the incremental Viewability does not produce the appropriate return.
Does higher Viewability drive performance?
Of course this isn’t a straight forward question to answer, as what constitutes ‘performance’ is different from one advertiser to the next.
If your primary goal is to make sure your ads are seen, then Viewability is the primary metric to measure against of course. But what about if, like most advertisers, you are trying to drive some form of conversion and return on investment?
Considering the above point about cost and Viewability, how should you be using this metric?
Let’s create an example using some basic figures. You buy 100,000 impression on three different ad formats, each with different Viewability scores.
This is a very rudimentary example as it assumes the viewable impression directly correlate with a share of sales, however it illustrates a point. When faced with this situation and the incremental cost difference represented it would still be more cost effective to by the less viewable inventory and absorb the wastage.
First things first. Measure it, report on it and understand it for your own campaigns. Until you do so you should not be making any decisions about how to implement it into your strategy.
Secondly I would recommend a period of testing. Understand the implications of only buying highly viewable inventory, and conversely less viewable inventory and what it means to your success metrics.
And thirdly I would recommend developing your own Viewability standards and benchmarks. Third party verification software such as Integral Ad Science and MOAT allow you to build you own definition of Viewability outside of the IAB standards. Once you have this set your own benchmarks and measure your activity against them to drive greater performance.