4A's Examine Media Measurement Priorities

The push for an industry standard cross-platform measurement is not only gaining momentum, it is also consolidating efforts across cooperating media entities.

As part of Advertising Week, the American Association of Advertising Agencies hosted a panel titled “Media Measurement Priorities” that covered the joint efforts of leading industry entities to facilitate cross-platform measurement and to decide, as an industry, what media measurement needs to look like in this new media environment.

“What we have now really doesn’t fit the bill,” noted Louis Jones, executive vice president, media and data, 4A’s.  “We need to have a collaborative point of view,” that also takes into account the needs of agencies.

From there, the 4As set out to coordinate the efforts of companies and organizations working on the issue and published a whitepaper titled "Media Measurement Priorities” as the first salvo.  



The paper set the stage for discussion of the most important priorities from an agency’s perspective. The top five were: unduplicated reach; currency; short term versus long term; walled garden and identity graphs; and attribution.

Agency Perspective

Even for these top priorities, there may be flexibility in the solution. Take, for example, currency. Historically, the TV industry has transacted on a strict set of metrics for currency.

Yet, for Jonathan Steuer, Chief Research Officer, Omnicom Media Group, “We are in a world that is complicated enough that if everyone had access to the right underlying data, different partners could agree to trade on different metrics and that would be okay.”

For Ed Gaffney, managing partner, director of implementation research and marketplace analytics, GroupM, the currency just has to be well understood, transparent and stable. “We can have multiple currencies,” he explained. “We have them now,” with digital and TV and even within TV there are a range of metrics. “As long as everyone knows how they are counted, and can use that data, for both sellers and buyers, it works well.”

For Gaffney, unduplicated reach is critical to address waste. But the barrier, according to Steuer, is that the measurement currency for TV “is based on volumetrics and not real humans” and is delivered, “on the aggregate and not the individual. We need a census to tie together and understand device delivery to actual humans.”

Industry Perspective

Other organizations besides agencies are deeply involved in the measurement discussion. George Ivie, Executive Director Media Ratings Council, explained the MRC’s two-year effort resulting in a brand new industry standard for video that was just released in early September.

This standard provides the framework for equalizing the exposures across platforms and de-duplicating it across some general principles.

The reaction from the industry was both accepting and yet guarded. Radha Subramanyam, Chief Research and Analytics Officer, CBS, noted that measuring, “viewability is a good thing. Nobody wants invalid traffic. Duration is important. But the devil is in the details. Implementation versus theory -- there is a big gap there.”

If you ask me, an effort that has created the foundation for the trans-corporate cooperation today has been through CIMM. This organization has been working on universal content labeling to help stitch together content on various platforms and devices through Ad-ID and EIDR.

Without a UPC-like code, there is no industrywide way to insure that content is accurately being captured wherever it airs. "It is an evolving time in television and we don’t have a granular, nationally representative impressions-based TV measurement system in place right now,”  because the data is siloed behind walled gardens and not shared, explained Jane Clarke, CEO and managing director, CIMM.

But, as there is strength in numbers, the first powerful step has now been taken. “The tech environment innovates. Technology improves. The standard is a first step in a long journey,” Ivie concluded.

7 comments about "4A's Examine Media Measurement Priorities".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, October 2, 2019 at 8:31 a.m.

    In addition to having  reservations about the value of what has been proposed as a "standard"---or, more correctly, as the beginning of a process that may, eventually, lead us to a meaningful "standard"--- I find it significant that there is no mention of other media--radio, print, OOH---in any of these deliberations. I also find it odd that representatives of those media are silent about being thus cast aside as being of lesser importance.

    Returning to the issue at hand, it's all very good to point out that we will get to questions like was the ad actually seen and what impact did it have later, however, I continue to harbor the suspicion that many will simply take the current proposal and run with it as if it  effectively deals with these vital factors. As has been pointed out---it doesn't. As I've said before," Haste makes waste". Eventually, we will have to face the fact that it will be up to each advertiser to decide how to evaluate the various media options---probably subjectively, but supported by some evidence--- as opposed to relying on some magical industry standard that saves them the trouble. Over the years I have worked with and for many advertisers but I have never met that illusory bird---the average advertiser---which is what general formulations seek to serve.

  2. John Grono from GAP Research, October 2, 2019 at 9:21 a.m.

    Interestng post Charlene.

    While I get Omncom's POV or other metrics (basically TV), but I see them as complementary to the 'currency' numbers.   There has been a plethora of complementary metrics over the decades so this isn't new.

    However, given the (unfunded) push by bodies like the 4A's for a single cross-media measurement system ... which metrics do we use?

    Or does Omnicom use their metrics, while WPP uses theirs, broadcasters use theirs, etc.

    It's hard enough fusing a handful of media currencies, let alone having other metrics.

  3. Joshua Chasin from VideoAmp, October 2, 2019 at 11:57 a.m.

    A couple of things.

    1. I think Jonathan is right about lmultiple currencies. Keep in mind that we live in a world with multiple currencies already; two of us can transact a deal in dollars, a different two in Euros, a different two in bitcoin. Clusters of buyers and sellers will be able to convene around different currencies, so long as buyers and sellers agree on the common currency in their transaction. (It will also be helpful if we are able to convert across currencies; how many Euros is a dollar? But that's a next level of complexity.)

    2. As regards other media beyond TV and digital (responding to Ed's comment here): I think when we as an industry conceptualize "cross-platform," there are two different phenomena going on that the term describes. One is the convergence of what has been TV, with digital; two things smashing together. The other i the fragmentation of what has been TV into different distribution channels, including both traditional linear, and digital:linear TV, VOD, SVOD, OTT, online video, etc. One thing breaking into lots of things. Increasingly, I think the cross-platform critical mass is around the latter; following premium video wherever it goes, and reporting on same in a holistic, unduplicated fashion. THis is not to say other media are not important; but advertisers have been spending on TV, radio, print and OOH for years without the industry grinding to a hault. Rather, I suspect the current existential crisis we are all grappling with is the measurement of what was once TV, and is now cross-platform video with multiple distribution platforms and access points. That appears to be the thing we need to gifure out, now. Once we do-- once we reinvent TV measurement as cross-platform video (or more aptly, coommercial) measurement, we can go back to the more mundane problems of determining how to optimize zcross TV/vdeo, print, audio, display, and OOH.

    3. One important trend worth noting is that TV companies (both networks and station groups) are moving from selling "ratings" to selling impressions. THis is inevitable and logical; it enables buyers and sellers to include all ad exposures, not just the ones that count in the archaic C3 construct. Migration to impression-based selling will go hand-in-hand with the rapid development of cross-platform measurement, because the inudtry as a whole is in the middle of migrating to a more addressable, targeted future.

  4. Ed Papazian from Media Dynamics Inc, October 2, 2019 at 12:16 p.m.

    Josh, regarding TV moving to "Impressions" this is simply expressing rating points in whole number terms---nothing new and always the basis for CPMs.

    As I have been pointing out recently, impressions across platforms are OK---assuming that they are truly comparable---for direct marketers who can relate their total impressions to actual responses and sales to determine if what they are doing is effective. However, branding advertisers are not direct marketers---nor are many of them going to convert into DR anytime soon. They are concerned primarily with the penetration of their campaigns against targeted groups and, of course, the frequency of contact---or, I should say, opportunity to contact as we have no measurments of commercial viewing. Hence, a media planner working on a branding campaign is not going to say, "get me so many impressions in late night TV in Podunk and so many in some other city, this tells him/her nothing about the anticipated reach and frequency that is desired. Moreover, a media planner ---or many media planners--are not going to allow a TV time seller to bundle TV impressions with digital Impressions---even for the identical ads--- without questioning the relative value of each type of "exposure". The stations will meet a lot of resistance on this score and will have to report the two kinds of "impressions" separately so the buyers and planners, in concert with their clients, can decide whether to equate them--or not. Also, I see nothing of this kind developing in national TV---just local, so far----and for very good reason. If any advertiser really wants to count "audience exposures" delivered by a seller, and include digital these can be expressed most easily and understandably as GRPs, not "impressions".

  5. Joshua Chasin from VideoAmp replied, October 2, 2019 at 3:52 p.m.

    Ed--the migration to impression-based selling is, I believe, simply a way to include ALL the impressions the seller has available as inventory, not just the ones counted in a C3. If moving from ratings to impressions was merely a matter of no longer using a denominator, then it wouldn't be news. But increasingly, media sellers are realizing that a smaller and smaller share of the impressions they generate are included in the C3 coin of the realm. Remove the denominator, but include EVERYthing in the numerator.

    (Parenthetically-- note the parentheses-- I've long pointed out to digital people questioning whether they needed GRPs, that they already have impressions; and if they can divide by population and multiply by 100, then a priori they have GRPs already.)

    Also-- impressions measurement and selling can absolutely support target reach and frequency. Indeed the cross-platfomr tools emerging today (from my and competitive companies) are engineered to provide exactly that. Planners and buyers will be able to know precisely how many different auto intenders or frequent fast food visitors or independent voters or domestic beer drinkers they reached in Podunk, and they'll be able to generate a frequency distribution against this reach (how many in-target were exposed 1X, 2X, etc.) precisely because we are measuring at the impression level. 

    As far as planners hesitating to intermingle TV and digital ads, what I see going on is an increased blurring between what is TV and what is digital. My family has a small place on Long Island where we've recently moved from cable TV to YouTubeTV. Even when my mom comes over and we watch Jeapordy on WABC, we're streaming it from a Google platform. Is that TV or digital-- or is it cross-platform? It's a broadcast TV station, and I'm watching it on the big screen in the living room; yet I'm streaming the content (and the ads) via Google.

    I don't know if that gets counted yet in C3. But clearly it should be.

    This is the challenge with cross-platform; we can't really think in terms of TV over here and digital over there. We need to think in terms of holistic and comprehensive measurement.

    Finally, on the subject of comparability, I tend to come out in favor of discrete line items in a measurement reporting tool, with each platform and creative length reported discretely. THen the user can decide how to treat the different line items (this also addresses duration weighting, assuming that one of the metrics is average second viewing for the spot length/platform.) 

  6. Ed Papazian from Media Dynamics Inc, October 2, 2019 at 5:07 p.m.

    Josh, I think that we are basically in agreement. However, I suspect that the spot TV sellers may not be planning to report the digital and "TV" impressions sepatately just as Nielsen, on a national basis, is melding out-of-home "audiences" with in-home. This goes for the original buys---as promised ---and also in the post buy phase where makegoods may be involved.Smart agency buyers will not allow this---if they are paying attention---and I hope that they are.

    Getting back to the issue of rating points or "impressions", consider the way this would work for a media planner for a branding campaign who is recommending a spot TV efort in 60 markets. Some of these may be high priority areas which get more weight than others, so the GRP goals may vary in many cities.Consequently,  In some markets the planner wants 150 18-49 GRPs, in others the planner wants 100 or 75GRPs  per week. Also, the GRP goals may differ by daypart for reach attainment  relative to CPM purposes. So in some markets the weekly GRP goal is 75 daytime GRPs plus 50 late night GRPs; in others, it may split 50 daytime/25 late night, etc. In order to convey this to the buyers the planner must go to the tedious process of taking the 18-49 population in each market and express the goals in "impression" terms, so the buyers know how many to purchase in each daypart. The planner must also notify the buyer of an adjustment factor to use----if any--- for the digital portion of the total impressions bought. After the buy goes down all of the "impressions" must be recalculated into GRPs to be evaluated meaningfully---another chore. Finally, when it's time to negotiate makegoods,if any, these too, must be converted into GRPs to determine if the desired goals were met in terms of actual audience tonnage. No brand manager is going to understand  that the spot buy went as expected if the results are expressed as "impressions" because each market has a different population base. More needless work. 

    If the sellers wish to get credit for their small but potentially valuable digital audiences, that's fine with me and I support it. But transparency is needed with both types of "audience delivery" reported separately. And please keep it simple. Report "TV" and digital audiences as GRPs and make everyone's lives less complicated.

  7. John Grono from GAP Research, October 2, 2019 at 5:08 p.m.

    Josh, I tend to agree that 'video' is the spark, but it is definitely not Robinson Crusoe.   Print publishers here are including digital + ink on paper audiences as their lead metric (EMMA).   Audio is facing the same challenges with streaming and podcasts.

    With TV ratings here, OzTAM does a +7 on broadcast (at a programme level - ad-breaks stalled), but they also have a VPM service for catch-up TV streaming.   So, our "TV set" coverage is pretty good and on a parity basis.   Next step is to include non-TV devices.

    OOH is worrking on their next generatiion of digital sihns plus wider scope.

    So at the end of those trails, we'll have pretty inclusive coverage of the exposures.

    Then we have to work out how to fuse them together.   Fusing data sets within each of those pillars is hard, and it's only going to get harder.   I tend to agree that the accepted measurement metric should be bespoke to each medium.   For example, time-based average audience is the norm for both TV and radio (and increasingly for online) - but should the thersholds be the same?   And if you consider print and press a time-based average would basically misleading (a person may read a monthly magazine diligently for an hour but over the month their average daily reading would be 2 minutes).

    Great discussion.   It's fun being on the bleeding-edge!

Next story loading loading..