2018: Will The Ad Industry Stand The Test Of Time?

Newsflash: The length of time an ad remains in view is directly tied to the success of the campaign. That might sound obvious, and more than a few journalists have echoed that sentiment whenever the latest time-in-view study hits the market, but understanding the real impact of time in view is complex. 

Let’s agree that time is money. And with recent studies showing that the longer an ad is in view, the higher the performance, wouldn’t it stand to reason that as an industry we rethink how we buy ads, measure their effectiveness, and reward all the publishers that contributed to the results? Still, we continue to adhere to a standard that tacitly implies that an ad that remains in view for two seconds has the same impact as one that was in view for 15 seconds, or more. Fundamentally, nobody would agree with this. Yet our adtech ecosystem recognizes those two time-in-view increments in the same way.



No wonder P&G, among others, called out the fact that “two seconds is not long enough” for an ad to be in view and the need for a single and better viewability standard, according to P&G Chief Brand Officer Marc Pritchard. It’s been almost a year since that public declaration at the IAB Annual Leadership Meeting last January. And those paying close attention will recall that the consumer giant gave its agencies one year to get to “a transparent, clean and productive media supply chain.” 

This year’s annual IAB event is right around the corner, being held in February, and there’s no doubt that we’ll all revisit the topic of viewability and hopefully duration of viewability as well. As an industry, we’ve got to get better at the time-in-view and viewability metric. 2018 is the year it needs to happen. 

Yet there are three gating factors at play. First, there’s the meager one second minimum viewability standard set by the IAB. Blink twice – try it – that’s a validated viewable impression. When the bar is that low, it enables a subset of less than desirable adtech players to subsist.

Second, the buying process is fragmented at best, broken at worst. In response, many publishers are bringing operations in house with mixed results. Let’s face it, what buyers really want is a centralized way to access quality inventory that drives engagement with ads that remain in view for a reasonable amount of time. 

Third is the issue of quality. No matter who you talk to, quality is a priority. Quality in terms of audiences, engagement, inventory, metrics, and the list goes on. When it comes to assigning quality to time-in-view, many publishers are ignoring the IAB metric and setting higher standards for vendors that want their business. That’s a good thing.

As an industry, where do we go from here? 

It starts with rethinking how time-in-view is prioritized, measured and rewarded. For example, the recent Magna and IPG Media Lab study found in one case that ads that were in view for four seconds increased conversions by 97%. Given this data, along with the many other time-in-view studies, it only makes sense that we start moving toward making the buys based on increments of time. 

Specifically, tie performance and pay to the response yielded by ad units that remain in view for a set period of time, say 10 seconds or 15 seconds. This isn’t an original idea, it dates back to the invention of the television ad. But in the online age, we’ve leapfrogged that old school TV model with more sophisticated ways of measuring engagement against time-in-view. What’s old is new again.

We also need to establish new and more realistic viewability metrics and hold all parties accountable to achieving, if not exceeding, them. This calls for a different attribution model. One that tracks performance and separates viewability (being seen) and time-in-view (how long the ad is seen). Armed with this data, buyers can make smarter decisions and drive better ROAS and engagement.

Finally, a broad and transparent supply chain, one where only high-quality inventory that remains in view for a set amount of time, would allow the entire ecosystem to better monetize its assets. 

For publishers to be successful in 2018, they will need to show the inherent quality of their sites and their users with more viewable formats and longer time-in-view. Better ads, not more ads. Otherwise, we risk eroding our brands, audiences, and bottom lines.

4 comments about "2018: Will The Ad Industry Stand The Test Of Time?".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, January 16, 2018 at 8:13 a.m.

    Exactly, Eric. However using a more ad-relevant metric goes counter to the sellers' interests and there are a lot of people buying digital ads who aren't very savvy about the numbers. Also, in order to put a "value" weighting factor on time-on-screen one must be willing to use human based metrics--like verified ad recall----into play and, so far, the digital medium remains wedded to using electronic indicators which can't do the job all by themselves. Until this changes, we will talk about time-on-screen endlessly, but not be able to solve the riddle. We cover this subject---cross media measurement---in some detail in our just released annual, "TV Dimensions 2018", and will delve into it again in our upcoming annual, "Intermedia Dimensions 2018", for those who are interested.

  2. Douglas Ferguson from College of Charleston, January 16, 2018 at 10:46 a.m.

    Good luck assigning a quantitative measure to quality. It's in the eye of the beholder.

  3. Jim Meskauskas from Media Darwin, Inc., January 17, 2018 at 11:03 a.m.

    Interesting thoughts. For a very long time I’ve discussed the notion of “time spent” metrics being used as a basis for both effectiveness and pricing (there are maybe half-a-dozen Old ClickZ and iMedia articles I've written on the on the subject). The challenge’s are these: 1) machine-based determination of time-in-view vs human. No small part of the problem with viewability is that we are relying on machines to determine a successful ad served, which is rife with error and opportunity for corruption; 2) the effort to determine pricing and cost-benefit against the time-in-view. If twice as long code twice as much, is it twice as effective? 3) just making sure the darn thing is really being shown at all still needs to be solved.

  4. Julien HIRTH from SCIBIDS TECHNOLOGY replied, January 17, 2018 at 4:47 p.m.

    Regarding point 2., for sure it would be totally sub-optimal for overall performance to buy only ads with a significant time-in-view because of the higher cost of those ads (the exact same thing for the viewability rate : V-CPA is generally better with a 60% V-rate than a 100% V-rate !)

    Actually it is the job of the bidding algo to assess if the time-in-view is a discriminative variable and if so adjust the bid for each impression according both to the conversion rate uplift and the market price increase

    With DSPs becoming programmable, it is actually quite easy to leverage that data (if it exists) into an algo !

Next story loading loading..