Despite the emergence of a Golden Age of data and ad tech, the ad industry still has fundamentally valued media either on an “opportunity-to-see” or a performance basis.
With the exception of some out-of-home media, which are measured on the basis of a propensity to see an ad, media buys have essentially been a binary process of “impressions” and “outcomes.” That’s about to change as several key developments begin shifting the industry to measuring what has actually been seen or heard, and perhaps most importantly, what people actually paid attention to.
The Media Rating Council already has put new standards in place that will enable audience measurement suppliers to create systems measuring not just how long an ad was in view, but how long consumers were exposed to it. Over the next year, it will focus on developing standards enabling those ad exposures to outcomes, effectively closing the loop between cause and effect.
But an interesting new metric has also taken root that the ad industry should actually be aspiring to fulfill: attention.
Like previous measurement buzzwords of the past -- especially engagement -- attention is still a little in the eye of the beholder. That’s why a group of attention-based measurement providers created The Attention Council. The council, which was formally announced during “Quality Metrics: Beyond Viewability,” a summit organized by Adelaide.
The summit, which features a keynote conversation with former Fox ad sales chief/now CEO of Attention Capital Joe Marchese, and panels with leading thinkers on the subject.
The effort is a good step in the right direction, but without a cross-industry authority like the MRC, or at least a big supply-side backer like the ANA’s, 4A’s and IAB’s 3MS, it’s not likely to coalesce as industry standards.
But that’s okay, because based on the work, I’ve already begun seeing some smart agency traders developing based on attention data suppliers like Adelaide, the evolution will happen on a grassroots basis.
Either way, the emergence of genuine attention metrics could finally be the answer to what Marchese has long identified as the ad industry’s equivalent of a subprime trading crisis.
Joe, the obvious next step in TV ratings---on a national level---is to add an "eyes-on-screen"---measurement in the TVision manner or--- something similar. This would require a parallel panel used in conjunction with Nielsen's audience ratings---or an entirely new service that can supply both audience and attentiveness information from a single source.
It's feasible right now---but the main obstacle is not technical issues that remain unsolved, but, simply, funding. Since attentiveness, however defined, is going to be a subset of the audience ---meaning that, on average, only 25-50% of the audience will be fully attentive per ad message---depending on the methodology----this will require a much larger sample than Nielsen is now using in order to get show by show, ad by ad ratings. So who is going to pay for it? Certainly not the advertisers and only some by the agencies. So the bulk of the funding will come from the sellers. Is it realistic to expect that TV ad sellers, who are adverse, quite naturally, to anything that is negative about their medium, will pay for a service that shows that many people watch shows but do not attend ads, even whan they don't zap them.
I have been the champion of TV attentiveness studies since ancient times and I was the one who introduced such measurements for TV shows in the early Simmons studies way back then. I still harp on this subject constantly in my Media Dynamics Inc publications. But I fear that until the networks are sold in as the likely primary funders---and accept that----we may be kidding ourselves. It's not just a technical issue---it's a business proposition that should be approached in that manner. What's affordable and what reasonable compromises must be made to get it funded---at least at the outset.
So you are saying they haven't been paying attention.
Joe - this is a really important topic and I've done some work in this field in terms of looking very hard at the research. Probably the best methodology for truly establishing valid measures of attention comes from Karen Nelson-Field of Amplified Intelligence (she was previously at Eherenberg-Bass Institute). Her learnings, combined with some strong work from RealEyes and interesting entities like Avocet are establishing a playbook that frankly could go far beyond the Nielsen paradigm.
Not to contest or diminish Sarah's comment, but we must remember that what is being asked for is a type of measurement that could provide reasonably definitive findings on a show by show basis day after day---the average TV show's average minute rating is something like two tenths of a percentage point----for all major and many minor TV channels on a national basis. Even if individual commercial airings could not be broken out, the average commercial in each telecast would be an acceptable substitute---meaning that each ad was measured but the results were aggregated. Getting this volume and type of information on an ongoing basis would, no doubt, require a panel type operation with a sample of about 70,000 households and 175,000 people. Other methodologies, such as those which expose people artificially to ads and invesitage how their minds respond, while interesting and sometimes more informative than simple eyes-on- screen indicators---invariably utilize very, very small samples and can not possibly supply the vast amount of readings reguired for a major update of the existing TV rating system. That doesn't mean that such research has no place in our thinking---it does---but mainly on a case by case basis and mostly in the development of the ad phase as opposed to trying to link it's impact with the varying effects of program environment, commercial clutter, commercial length, time of day, demographics. etc.
As I noted in my first post, this is not merely a theoretical matter nor does it call for a perfect answer----if such a thing exists. To get something meaningful done, we must select a metric that is readily utilized and one whose findings are easily understood by all parties---sellers and buyers. And the measurement must be compatible with the practical requirements of supplying ongoing information, at reasonable scale and cost---in order to be a viable business proposition.
Great discussion.
To pick up Ed's comment about "eyes-on" (i.e. viewing rather than presence) it should apply to TV, digital and OOH.
In Australia, for OOH, we use "likelihood-to-see" (LTS) and not "opportunity-to-see" (OTS). For some signs LTS is virtually the same as OTS - think Times Square. For other smaller formats LTS may be one-third that of OTS. This is based on a 'point-in-time' study of formats - not the actual ad on the OOH faces.
For TV I would like some 'eye-gaze' data. For example, it could be from the TV panel (not my preference) or a 'side-panel' e.g. households that have recently left the panel. This would measure how they watch rather than what they watch. Again it would not measure individual ads but would be a valuable improvement.
The elephant in the room is eyes-on for digital ads We still pay disproportionate attention (no pun intended) to what is being served and not what is being seen. For example, I have just one browser open and just three tabs open in that browser. Clearly my focus is this Mediapost tab. But what about the ads being served on the other tabs getting a +1 impression count? I know that in my usage patterns there would be many more orders of magnitude of unseen ads on my computer than my TV.
Sarah raises it to the next level - Karen Nelson-Field. (Yes, another great Aussie media researcher - read Byron Sharp's How Brands Grow if you haven't already). Karen's research focuses on the attention that people pay to ads. NeuroInsights here in AU is also doing great work on brain response to ads.
I see the approach to gauge 'eyes-on' LTS as opposed to OTS as very important and would love to see it as part of the media trading currency. The networks etc. could still spruik their content on the 'big number' (e.g. average minute audience and even episode reach) to the public, but trade on this sharper number. Of course this raises the issue of the programme having a higher LTS than the ad's LTS - good luck looking for make-goods if you make a bad ad!
But crucially we also need LTS for online video if we ever want to get an 'all video' measurement system, because you can NOT have two different measurement systems in a 'unified' currency. (Some forms of multivariate data modelling can differentiate effectiveness, but that is post-event and no use for ground-zero strategic and implementation planning).
I see the Nelson-Field approach more as being a brand effectiveness measure that sits on top of the unified LTS currency. And yes the brand needs to fund it as it would be unique to the brand and because the variability of effectiveness between ads is likely to be much more than for a programme.
If the industry decides to use "Attention" as a surrogate for Ad Exposure, Eyes-On or Visibility Adjusted Contact or VAC and nothing more (very important!) and not to be confused with "impressions', then in parallel with John and I believe Ed, the industry would have a universal, harmonised comparable media currency.
Per the ARF's, "Making Better Media Decisions", engagement, impact, brand awareness, propensity-to-purchase, and ultimately sales are all driven from the base of achieving ad exposure but are not media measures as such as they involve so many other elements especially creative, still the driving force of any campaign.
Circulation, traffic or served, opportunities-to-see and likelihood-to-see are each very different in value and are the subsequent stepping stones to achieving the goal of ad exposure for which media platforms take responsibility and without which none of the subsequent Ad & Campaign effectiveness measures can be achieved. (My article in Media Post , "When is an Impression not an Impressiosn" may help understand the fundamental differences.)
As we always asked at the media agency when presented with any media data, "What is the defintion of "audience" and what is the defintion of "impressions"? Those two generic media terms can cover a great many sins and typically require significant adjustments to put the data from different media platforms on a level playing field for meaningful comparison.
It should be noted that OOH globally has generally embraced Visibility Adjusted Contact or VAC which is a measure of ad exposure or Eyes-On and as such does not require adjustement in any media mix evaluation. The use of PPM to measure radio in major markets also reflects a measure of ad exposure but based on being reported at the average commercial minute level.
Look forward to Media Post's Summit on, "Achieving Universal Currency Across Media Platforms". And by-the-way should that currency include duration weighting?
Tony, of course, I agree that a physical manifestation of "attentiveness" such as eyes-on-screen is not only the best way to start but it may be the only feasible way to get the huge amount of data that will be needed. The problem is how to define the "audience"---is it everyone in the room for TV? Or is it those who were in the room and "watching" the program for a period of time before the break? And how do we define attentiveness" to ads using "eyes-on-screen" technology. Is average second the best way?Or is some sort of time span indicator needed---like eyes-on-screen for at least two -thirds of the commercial's length? As regards OOH or digital, more questions are raised, among them, how do you measure eyes-on-screen" for ads on smartphones and 30 sheet outdoor posters?
Eyes-on-screen seems to me to be the best and most easily understood metric that might gatrner universal acceptance----but even so, a lot of work needs to be done if we are to move meaningfully in this direction.
Ed:
I would posit that OOH (excluding Digital Place-based network measurement at least in the US) is there in terms of Eyes-on-Display (or on screen or panel) and has been for some time. OOH has also embraced the fundamental requirement to be in the visibility zone with eyes-on the display to count as "audience", a broad generic term, which OOH determined was better described as "Visibility Adjusted Contact", VAC or "contact" for short. In other words a highly refined and defined "impression".
You raise many valid considerations regarding how to apply these eyes-on-screen priniciples, which should include the visibility zone requirement, to other media platforms to produce a common harmonized cross-media currency. However in your example of ads on smartphones versus 30 sheet posters or even versus cinema I am not sure it needs to be that complex.
Media agencies have been dealing with assessing and comparing screen or display or page size or ad duration for ever along with the myriad of other campign and media oriented elements that will make the ad campaign more relevant and effective, i.e., impactful. We used to develop a composite adjustment by media platform and often by media vehicle for modifying ad exposure or eyes-on-screen data, our common currency, to impacts for any campaign for any brand target group. As you would fully understand such an impact weight was the result of co-mingling many elements based wherever possible on category research and/or experience and good judgement. What we did understand was that impact weights could only be applied to audiences that had been exposed to the ad in each medium versus applying it to some mix of divergent defintions of various so called audience impressions. Without a known ad exposure or eyes-on-screen or contact there can be no impacts of course.
There will be winners and losers if all media base currencies were eyes-on-screen within the visibility zone (or visibility adjusted contacts). It would surely be the greatest possible service to advertisers and their media agencies? So will CIMM/ARF and/or MRC embrace this fundmental media measurement foundation for a common basic common currency any time soon?
Cannot wait for Joe to set the date for the Media Post Summit on this topic!