“The Sell Sider” is a column written by the sell side of the digital media community.
Today's column is written by Aryeh Lebeau, executive vice president of client operations at Remedy Health Media.
A wise man once mused: “If an ad loads on a web page but no human sees it, does it count as an impression?” And so, we in the digital media world embarked on the wonderful, wild ride of measuring viewability.
While this key metric in our industry is old news at this point, the ramifications of optimizing toward and especially billing on viewability are worth exploring.
Overall, the advent of viewability has been a net positive. Any time digital media can become more sophisticated is a plus in my book. Viewability has been an obvious win for advertisers who benefit from far less waste in their buys.
On the publisher side, despite a painful transition period, viewability has had significant upside. It’s an opportunity for publishers that have adapted effectively to gain a competitive advantage in the marketplace. Strong viewability is a draw for repeat business and justifies increased rates.
Viewability, however, is not an end unto itself.
The danger of single-mindedly focusing on viewability is that in some respect it may be antithetical to performance goals, and other KPIs may suffer. For example, publishers are often forced to optimize away from larger, attention-grabbing ad units which tend to perform best simply because they are tougher to keep in view; this is particularly true when the definition is 100% of pixels in view.
The MRC And IAB Got It Right On Billing
The goal of 100% viewability remains an unreasonable expectation, as previously stated in the IAB’s 2015 Transaction Principles, despite continued improvement in measurement and best efforts on the publisher side. While billing solely on impressions that see the light of day is a noble cause, it’s only fair to build in a buffer to account for numerous factors that are out of publishers’ control.
Advertisers, for example, share some responsibility in building creatives that load efficiently. I’ve seen campaigns that dropped as much as 15 percentage points in viewability overnight after a creative swap. Although they may run on the exact same pages and devices, the campaigns took a huge hit simply because creatives took seconds longer to render.
The end user is also a significant variable out of the publishers’ control. There’s always that one guy who likes to keep his browser at a 200% zoom, pushing the ads out of view. It’s also hard to account for connection speeds, particularly on mobile, which can produce wide variation in viewability rates.
And, of course, while measurement continues to improve, there are still sizeable discrepancies seen from one vendor to the next. These are just some of the reasons why 100% is not the way to go. A guarantee threshold, as recommended by the IAB, allows for a margin of error.
Not All Pixels Are Created Equal
The other night I was in bed watching TV when I noticed that my daughter’s hairspray on the dresser blocked a tiny portion of one corner of the screen. When the good folks at Arby’s aired a commercial, I was still left fully aware that they possess the meats, despite 3% of the screen being obscured.
Contrary to the 100%-of-pixels-in-view definition that some agencies have adopted, every single pixel of an ad does not need to be in view to provide value to the advertiser. Advertisers still gain significant benefit from ads that are 99%, 70% or even 51% in view. End users can still see most or all of the messaging, interact with the creative and click on it, yet per the stringent “fully on-screen” requirement, publishers do not get paid for this inventory. The notion that an ad with a single row of pixels off-screen is invalid is unreasonable.
Fourth-Party Billing: A Bridge Too Far
DoubleClick for Advertisers (DFA) and other ad servers have been the accepted source of billable impressions for quite some time and should remain as such. Some agencies have begun billing on viewable impression counts from verification vendors. This approach eliminates as much as 20% of DFA-recognized impressions off the bat because they were not analyzed by the verification partner.
As the MRC said a few years ago, “Nonmeasured impressions should not be assumed to be nonviewable impressions.” Doing so lays waste to the value of the publisher’s inventory unnecessarily.
The right way to handle this is to use the verification companies for their intended purpose: to measure viewability and monitor fraudulent activity. The viewability percentage should then be applied against DFA delivery for billing.
Ultimately, there’s no question that viewability should remain one of the core tenets of digital advertising. While there have been significant strides made over the last five years, there’s still room for improvement, particularly on the billing front.
Publishers and advertisers alike need to work together to hone the best possible practices and most effective industry standards.