| |

Dentsu: Forget ‘viewability’ and focus on attention in online media buying

Dentsu: Forget ‘viewability’ and focus on attention in online media buying

“Viewability” should no longer be regarded as a meaningful measure of advertising effectiveness, Dentsu has concluded in its latest research into online advertising effectiveness.

Instead, the owner of media agencies Carat and iProspect is arguing for “attention” to be used as a more reliable predictor of a how successful an online campaign will be.

Using attention as a key variable is a departure from conventional programmatic media buying, which is optimised according to audience targeting and data collected about internet users.

The conclusion is also a challenge to existing Media Ratings Council standards, which say 50% of an ad must be in view for one continuous second for display and mobile. Large display formats must have 30% in view for one continuous second, while video ads must be 50% in view for two continuous seconds.

Dentsu’s research says: “MRC Viewability Guidelines do not fully represent real audience behaviour and sticking to them verbatim may not improve ROI for all advertisers”.

This is because different formats can deliver outcomes such as choice and recall at varying levels of attention. For example, three seconds of eyes-on attention on one format is not the same experience for the user as three seconds on another.

Instead, the report suggests online media buyers should be looking at, “seconds of Attention per 000” metric, combined with data that informs how these seconds have the power to translate to outcomes for brands.

However, Wayne Blodwell, founder and CEO of The Programmatic Advisory consultancy, told Mediatel News that having an independently accredited standard for how viewability is calculated is important.

“Advertisers, agencies and publishers need to trust the methodologies viewability vendors use and this is difficult and time consuming to audit technically, so having the MRC as a resource for that adds value,” Blodwell said.

“Lots of buyers and sellers won’t use the recommended minimums the MRC suggest, they create custom metrics, but the technicalities that sit behind that need third-party auditing by MRC.”

The report also highlights purportedly more important variables for an online campaign’s success, such as whether views are forced versus optional, the role that sound can play, creative fit and targeting. None of these factors are captured in any current media value system, the paper adds.

The findings come nearly two years after Dentsu claimed a UK media first for using a custom Google algorithm to optimise a campaign for viewer attention, with the help of eye-tracking research from Lumen.

Optimising campaigns for attention may be an alternative digital media strategy as third-party cookies, used to track web users for marketing, fall out of favour.

Google is expected to block the use of third-party cookies on its Chrome browser, long after Apple’s Safari and Mozilla’s Firefox did the same in 2019.

Hamish Nicklin, CEO, Media, Dentsu UK&I, said: “At Dentsu we are serious about changing the way media is planned and bought for the better and we want to drive that evolution forward by offering tools powered by insights which deliver our clients meaningful outcomes at relevant value.

“This new research drives home the importance of the human experience of media and leaves us with no doubt that attention is a more reliable predictor of outcomes for brands than existing measures of success.”

BillHarvey, Chairman, RMT, on 31 May 2021
“Totally agree Philip.”
PhilipTownend, CGO, Amplified Intelligence, on 28 May 2021
“Bill couldn't agree more, however attention can and will be predictable based on advanced algorithmic models which can look at an impression and make a very accurate forecast informing bid and pricing strategies.”
BillHarvey, Chairman, RMT, on 28 May 2021
“Truly one can use more outcome predictive metrics for buy optimization than impressions or even viewable impressions. Before the buy is made one cannot use attention itself because one doesn't know what it will be until afterward, being affected not only by the unit and the context and the audience but by the creative itself. However one can use predicted attention (Adelaide) or averages by media/unit/audience type (TVision, Lumen, Realeyes, Amplified Intelligence) or resonance between the psychological content of the ad and the context (RMT) all of which will be more predictive of the outcome - or use combinations and random control trials to optimize the metrics used - likely to be more than one of the choices I've listed to get max incremental ROI and brand equity effects - worth testing experimentally and the investment is small for a large result.”

Media Jobs