Corporate Blog

Sense and Viewability

This blog post originally appeared on Turn was acquired by Amobee in April 2017.

In case you missed it, Procter & Gamble Chief Brand Officer Marc Pritchard gave the “biggest marketing speech in 20 years” at the IAB Annual Leadership Meeting, invoking marketers to stop putting up with shenanigans in digital media. Pritchard in part blamed the media supply chain, which he dubbed “murky at best and fraudulent at worst.”

Atop Pritchard’s list of remedies was a uniform viewability standard. That would be a positive step in alleviating confusion in the market. But, as Pritchard no doubt knows, the viewability issue is complex. A standard measurement will help, but viewability will always be a bugbear for marketers. Pursuing viewability at all costs can be counterproductive. The smartest approach is to accept somewhat lower visibility rates to ensure maximum reach.

Putting Viewability in its Place

While it’s true that an ad can’t be effective if it can’t be seen, it’s also true to say that an ad that can be seen can be ineffective. The ad might, for instance, be seen by the wrong audience. That’s one reason that viewability in itself is incomplete; just because an ad is viewable doesn’t mean that it will get a consumer’s attention.

One pragmatic approach to this issue is to establish an engagement metric that’s independent of viewability, that way you can see how engagement correlates to different viewability levels. Often there’s a point of diminishing returns for viewability and engagement.

Viewability also has no bearing on other variables that can boost an ad’s performance, like contextually appropriate environments, frequency and targeting. In other words, the ads might be viewed, but that’s beside the point since they’re not working.

The Goldilocks Approach to Viewability

There are a number of levers marketers can pull to maximize for viewability in such a way that will also increase ROI and campaign performance. First, to determine levels of viewability, marketers can go with private marketplaces and whitelists for the highest levels of viewability. They can also use pre-bid targeting to set viewability rates or incorporate viewability rates into the real-time optimization process.

Next, marketers can determine their own viewability standard, either by using the MRC’s standard (50% viewable for one full second for display and two seconds for video) or GroupM’s (100%). Then they need to figure out which percentage of viewable impressions to aim for (as mentioned, the IAB recommends 70%.) There are other factors as well, including whether the audio should be on and the minimum time a video ad needs to make an impact.

Once those variables are selected, the marketer can A/B test different levels of viewability to determine the percentage in which the link between viewability and engagement trails off.

What often happens is that there’s a sweet spot in which the viewability percentage tracks nicely with the number of impressions delivered, which nets a fairly low viewable CPM (vCPM). A campaign with low viewability means you’re paying too much for ads that aren’t being seen. A campaign set for high viewability means you’re paying a lot for a fairly low number of impressions. Determining the right number takes some testing.

The Long View

Viewability is just one variable within a campaign. Though it’s logical to assume that simply striving for very high viewability will lead to a more effective campaign, the data doesn’t support it. Getting the industry on board with a standard metric, as Pritchard proposes, is key.

For more on viewability, see our ebook, “Bringing Viewability Into Focus.”