Will 'Attention' Metrics Bring 'Impressions' Reform?

“Attention” measurement grabbed the limelight on the opening day of the ARF Audiencexscience Conference Tuesday, but it was Havas Media Group Global Head of Activation Jon Waite who may have raised the most crucial question, asserting that the term “impressions” has become a nebulous term in the audience measurement industry.

Lumen Research Managing Director Mike Follett underlined the value of attention metrics and TVision Insights Manager of Data & Analytics Kelsey Hanlon concurred, noting, “Attention is a quality metric that is predictive.”  

A panel led by Jon Watts, managing director the ARF’s Coalition for Innovative Media Measurement (CIMM), could not agree on whether attention should become a media currency in an evolving, potentially multiple currency marketplace, though it was that it has become an invaluable media planning tool.  

CBS Chief Research and Analytics Officer Radha Subramanyam prudently expressed caution on the dilemma of, “lumping media attributes with creative messages in relation to brand goals and lifecycle of an ad campaign.” 



Havas' Waite, together with Caroline Hugonenc, senior vice president-research & insights of Teads, reminded the ARF conference’s virtual audience that attention, whether for content or ads, was a precious resource for consumers and “does not happen in a vacuum.”

“Attention seconds per thousand viewable impressions,” or content rendered counts, vary widely by media platform, format, daypart, and program genre and engaging media content really matters.  Quality journalism counts, with news driving the highest engagement from its audiences along with attention to ads played in that environment.  

Dentsu Vice President-Director of Global Media Partnerships Joanne Leong outlined how the agency is bringing an understanding of attentive seconds by media platform, format, duration, context, brand, target group and creative message type to planning and buying as related to predicted outcomes.  The foundation being Cost-per-Thousand Effective Attentive Seconds, or CPMEAS.

Adelaide CEO Marc Guldimann and NBA Senior Director-Data Science Lead Erik Nylen suggested that attention offers a consistent measure across all media that is predictive of brand outcomes, and increasingly offers scalable depth that “impressions” do not have. 

To add understanding to the value relationship between creative attention and media qualities, Realeyes’s Max Kalehoff and TVision’s Tristan Webster fused creative pre-testing and in-market ad attention.  

Their findings underlined the value of media environment and the context of a programing/editorial related to attention for an ad.  In fact, a low performance ad can produce “good” attention scores if placed in premium environments, whereas high performing ads tend to wear out relatively quickly.  

In his keynote, producer and Fordham and New York University professor Evan Shapiro highlighted the “Trillion Dollar Death Stars” in the media universe that are driving the supply-side of the business: Amazon, Microsoft, Apple, Meta, and Alphabet.  

Using marvelous cartographic maps – the relative size of the “Death Stars” that touch every other business, even against major media corporations – he noted the key attributes to survival are: utility, consumer value, audience superstars, subscription versus ad revenue balance, influence on Gen Y, Z and A, diversity of products/services, and corporate flexibility or “flywheel.”  

As a company that, “Takes your privacy and sells it to others,” he appears cold on Meta’s future and considers Roku and Netflix as likely M&A targets in the near future. His stars: Amazon and Disney.

2 comments about "Will 'Attention' Metrics Bring 'Impressions' Reform?".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, April 13, 2022 at 1:08 p.m.

    Good review, Tony.

    I see that, as usual, the vitally important subject of attentiveness is being treated as a theoretical matter without regard to the opposing forces that are at play and the refusal of many CMOs to grasp that the way they buy national TV time---largely in CPM-driven futures ----or "upfront"--- negotiations, negates serious attempts to zero in on individual, brand targeting needs which are often mindset as well as demo/ product usage driven.

    Since the sellers will make the ultimate decision about whether attentiveness is part of the "new" TV rating system, it stands to reason that no matter what theories are floated, that attentiveness will not be the basic "audience " metric and that device usage based "impressions" will "win". Perhaps attentiveness will play a small role, here and there---usually in isolated cases where it's in a seller's interest--- but, sadly, that's about all that I expect to see---and I really hope that I'm wrong on this.

  2. John Grono from GAP Research, April 13, 2022 at 10:35 p.m.

    Thanks for the post Tony.

    I am far from impressed with the stampede towards 'impressions' as a media metric, basically because the 'definition' (if it actually exists) of what an impression is nebulous at best.

    Probably the closest metric is average ad break ratings.   Here in AU we didn't go down that road because (i) we have day-after access to minute by minute ratings and can calculate the audience reduction (I know Ed, many don't 'log out' during ad breaks), and (ii) the propensity to switch off/channels is very variable - you can calculate an 'all ads' degradation or with diligence by programme as a proxy.   It's hard work but it can set some usable benchmarks.

    But what I am hearing is that there is a big appetite for ad-attention.   I totally unerstand that.   But when you think about it, an ad-break may contain a half-a-dozen+ ads.   Each brand would want to know its ad-attention.

    Programme attention can be estimated reasonably well using the episode's reach and average audience.   The lower the ratio of reach to average audience indicates less channel changing, reflecting behavioural attention but not cognitive attention.

    But you can't apply the programme attention (reach/average audience) to the ad break, as the cognitive (rather than beahvioural) attention in ad-breaks (which is what the advertiser wants) is extremely volatile.

    Of course the brand could do some cognitive attention testing of their ad ... but they would have to pay for that.   Good luck getting that to happen.

Next story loading loading..