Over the years, I've covered a number of controversial shifts in the methods used to establish the value of media currency. But I’m confused about the one currently growing around the video marketplace’s shift to a duration-weighted impression.
I understand that people don’t like change, but the reality is that change is constant. The duration-weighted impression is an ingenious solution for managing a fundamental change that's already taken place in the marketplace: the fragmentation of video advertising and its distribution across screens.
The solution gives people on both sides of the marketplace a simple common denominator that both buyers and sellers can use to assign value to a video advertising impression. That’s all it is, a denomination.
It’s up to the marketplace to determine what the value of those impressions are, and that happens through a variety of processes that influence demand: research, perceptions, selling, etc.
In other words, the duration-weighted impression enables the ad industry to size the supply of video advertising inventory based on an equal common denominator, so that the supply and buy sides can haggle and influence each other about the values assigned.
It is not, as some might fear, a means of commoditizing video ad inventory. It’s just the denominator. The value comes from how it's numerated.
And that’s exactly how the Media Rating Council conceived it.
The next phase, says Media Rating Council CEO and Executive Director George Ivie, is to begin creating the means for the industry to fairly and equitably begin assigning value to those units of exchange.
If you ask me, that’s where the real controversy should begin -- not with the duration-weighted impression, or common denominator -- but how value is assigned to it, because that’s where the winners and losers will be determined.
In fact, one of the reasons why the MRC delayed the implementation of duration-weighted impressions until 2021 is that it hopes to have those value systems in place so that they can roll out concurrently.
I wish them good luck, because if creating a common denominator proved to be controversial, establishing a means of valuing it could prove even worse.
Ultimately, marketplaces find their own symmetry based on a combination of factors, including market data, research, and analytics. Beyond that, it’s all about perceived value, and that’s where selling and spin come in.
You are right in wishing "them" good luck, Joe.
I hope that everyone realizes that the duration weighting part of this equation is only a first step. Next----and very, very difficult to do---is figuring out how to place a fair indicator of probable ad exposure at the various weighted levels---like distinguishing between TV "15s" and "30s" and their various digital counterparts. Do you attempt to develop a separate "value" barometer for desktop computers, smartphones and tablets? Do you evaluate same screen ad clutter against full screen presentations? What about engaging vs nonengaging content? And demos? And what about print media, radio, out-of home billboards? Most difficult of all, what metric is appropriate against all media types and situations---verified ad recall? If so who will fund and/or do the humongus study that tries to do this very complicated job?Or, cutting it down to manageable size ( maybe ) are we concerned only with TV and digital video?If so is that fair?
It is in theory no different from equivalizing units against a base :30 as we do for linear video, that conventional stuff on which we we spend billions. It is not an engagement or effectiveness proxy, it is a a value per second metric which turns the comparison of apples to oranges into apples to crab apples.
More education about media metrics, impressions and currency will go a long way to soothing this unnecessary angst. DWI is actually just an adaptation of a solid and proven value-to-cost metric that is sorely needed.
I am afraid there are no “shoulds” when it comes
to media methodology, duration weighting and
Candidly, on its face, I find the notion of a “common denominator” in this context conceptually off-putting, if not wrong. If the numerator is composed of disparate components, what are we doing pretending that we can find common numbers as in grammar school math?
Now that I have begun to enjoy the benefits of SMART TV first hand and all its disparate offerings, I think it is prudent for the MRC to go slow at this juncture. Perhaps even slower. So what are we to do now?
I think Ed Papazian has asked a number of crucial questions that need answers now. He’s also generously offered crucial context for addressing
the looming controversies.
At the same time, I am taken with the practical wisdom of Dorothy Higgins whose words resonate with me: “More education about media metrics, impressions and currency will go a long way to soothing this unnecessary angst. DWI is actually just an adaptation of a solid and proven value-to-cost metric that is sorely needed.”
Yes, more education may enable us to answer the critical questions before us. But not all impressions are created equal. And there’s nothing common
about them. There never was. I’m sorry. We need
to stop “shoulding” all over ourselves as we address this matter professionally.
Nicholas P. Schiavone
@Nicholas Schiavone: All the duration-weighted impression is, is method for crediting that a video impression exists as piece of advertising inventory. The durations are weighted against a :30 because that has been the closest thing to an ad industry standard for video inventory. That's different than the value people assign to it, and I don't think there will be one single "barometer" for doing that. Whatever methods the industry ends up using for valuing those impressions, individuals in the marketplace will likely assign their own values on top of that, based on how much they demand it.
I don't think this is that different than the history of TV advertising. Yes there were rules of thumbs when :60s were split into :30s and :30s were split into :15s, etc., but there was never a single industry standard for crediting the impression value of those durations of inventory, and I don't think that's what the MRC is doing here. It's just an accounting principle so people have a common denominator for applying higher order values on top of it.
That said, in the history of TV, I can recall times when some individuals asserted that there were different values inhrent to the "screens" people were seeing them on. In the old days, the screens were differentiated as "broadcast," "cable," "satellite," etc. screens, but and the people asserting the differences in value among them uses all sorts of research and arts of persuasion to convince others.
That's how marketplaces work.
If you convince enough people perceptions and demand change accordingly.
Based on your comment, it would appear I am insufficiently evolved, if not completely mistaken in my thinking on the topic of duration-weighted impressions. You didn't even
give me credit for embracing the thoughts of Dorothy Higgins on the topic.
Your comment about the "old days" when people differentiated value based on "screens" implies that my thinking is that of a media researcher in the 1950s. That would be quite
What is really slick, is the jump from references to "accounting principles" to "arts of persuasion" where the media accounting and negotiating process is concerned.
Clearly, the soul of research and measurement are without an ethical or intellectual compass
in this potentially controversial affair. What do accounting principles say about counting monopoly money as secured currency? Perhaps a sophisticated weighting scheme could be worked out here, with an emphasis on sophistry, of course.
I have only one point. Duration-weighting may be the best we can do now, but I despair about the consequences of using anything short of integrity, intelligence and rigor in determining the so-called "value-systems" that will be needed next.
@Nicholas Schiavone: Maybe we're discussing different things. My understanding is it's just a means of crediting how much inventory there is, not what value it has.
I don't understand why that's controverisal or why people would dispair.
It's just a way for sizing how much of something exists. Not how much it's worth.
Perhaps we are discussing different things.
What is the "something" being sized?
It would seem that I need to review the definition or basic criteria for identifying "inventory." Is it comprehensive? Is it reasonable? Is it decipherable?
I do not expect you or MediaPost to have all the answers. You already perform a great service by providing sunlight as "the best disinfectant" for this issue, in the spirit of Louis Brandeis.
If you can offer any public direction on getting up to speed on the matter, I, for one, would appreciate it.
"Something" = cross-platform video ad impressions.
For your review:
In sweepstakes this is easy because the end date of the sweep determines the end of the campaign. At that is how is suppose to work. Often the excess inventory might run ads for the then expired sweep might run for another 30 days. That's a small problem.
In short there should be a real end date and stop time for all ads with a return of unused ad dollars.
I believe Joe has perhaps unknowingly helped us understand the serious confusion that has been created by the term "viewable impressions" in the proposed Cross-Media Audience Measurement Standards (a device measure and not an audience or more importantly an actual exposure or "contact" measure) and their potential duration weighting, when he states his, "understanding is it's just a means of crediting how much inventory there is, not what value it has." Really!!??
As readers of this piece and the comments please read "Duration Weighted Impressions: Is it worth it?" by Vas Backopoulos in Media Post, June 20th AND the comments.
I suggest it may further help resolve the confusion on this matter albeit, in my opinion, neither Dorothy Higgins, Nicholas Schiavone nor Ed Papazian are confused as they see this addition on Cross-Media Audience Measurement (Video) for its impossibility as a meaningful addition to the proposed Standards as proposed. As stated in the comments on the earlier article, "stick to the mandate of measuring exposure (or contact). Without it there can be no effectiveness of any ad!" BTW: Per John Gruno, Viewable Impressions (Video) are merely a measure of "served, rendered or completed as key reporting parameters." In traditional media such a measure is merely a "proof of media's content delivery to spec" and as such completely different from an audience gross-impression or contact metric.
Advertisers want everything tied to outcomes and response by brand and by campaign. Of course! The variables involved in understanding that are numerous, complex and interrelated. Viewable impressions (and Duration Weighting) cannot be forced to get you there. Time to re-review the ARF Making Better Media Decision Model?
* In the standard, ONLY the portion of impressions that are 100% viewable are credited in the duration weighting.
* Vas has a point of view. I have a different one.
* Outcomes are the next phase.
We need to understand what the 'responsibility' of the media owner is.
That then needs to be juxtaposed with the 'expectations' of the advertiser - which ranges from immediate sales, to sales-this-week, sales-this-month, sales-this-year to building the brand for the long term.
However, ultimately the media owner does not have the responsibility for sales or brand health. The media owner has no input, say or control over the strategy (thank goodness), the creative and the implementation. They do have control over delivering on the implementation - using TV as an example - the correct ad in the correct programme in the correct break (if negotiated) and the correct position-in-break (if negotiated). If there is a CPM agreement (and there should be) then the campaign needs to deliver on that.
But muddying the water are media owners/channels claiming correlation as causation.
For example, the online ad attributes success as it was the last touch-point before the sale. The TV ad may claim success as well because 3 minutes before the sale the user saw a TV ad, logged in and puchased. Meanwhile radio claims success because of the ad they ran in drive time to watch the TV programme that night. But the OOH billboard along with print/press set the campaign up with their teaser campaign.
I stress again that ONLY THE CLIENT (or its agency or media analysts with access to all date) has access to all this data.
What we MUST do is provide comparability of ad impressions within delivery. That is video needs to be comparable for TV, online and cinema. Audio needs to be comparable for radio, online, podcasting etc. Visual needs to be comparable for print and press and probably for OOH.
Comparability is yet to be decided. For example a minimum threshold of timed viewing for video. A minimum threshold of timed listening for audio. A minimum size and gaze time for printed/physical media.
Then within such parameters an agency or advertiser (i.e. the creative side) can decide on a campaign's success.
The difference with this approach to duration weighting for ads is that you either pass the threshold as a viewer, listener etc. rather than being a 'part viewer' etc.
I also re-state that currently in the digital ad-serving world I don't see a plethora of meta-data (e.g. total ad-duration) in order to do the duration weighting calculation to deliver 0.7341 of a viewer. I feel that the IAB's 2-second threshold is about right. It is simple, clear, uniform and easily deployable.
I am struggling to find one thing that duration weighting makes better. it is a horrible equivalency between ad units of different time duration as no MTA model ever gives relative effectiveness as magically being the same as reative duration. Media is not "4 quarters for a buck" type of equivalency although duration weighting seems to want it to work that way. 15"-s are not half as effective as 30"-s on TV. 6"-s are not 40% as effective as 15"-s Pricing will reflect that and units will NOT be priced proportional to time. It also screws up the reach and frequency curves...now we will get fractional calculations that differ from nominal ones and it is likely that Duration weighted reach will actually make reach curves MORE discrepant with sales outcomes. So what problem is duration weighting solving? I am at a loss.
Unfortunately, most advertiser and media people I talk seem to think that duration weighting may be the magic solution to determining "value" for ad "exposures" across media platforms. It's not. No matter what calculations are used each advertiser and each ad agency is still going to be required to make judgements----hopefully supported by some research evidence-----when making cross platform ad exposure and impact evaluations. This applies even within the same medium---like comparing the value of TV "15s" with "30s".
Lots of good comments and reactions.
Couple of replies here.
@Joel Rubinson: I'm not sure "better" is the right question. The duration-weigthted impressions are something new: a common denominator for crediting when -- and for how long -- a digital ad impression is counted.
I think the problem with some of the comments in this string is that people are expecting it to do more or be something else. And I also think that is where some of the "controversy," fear and spin are coming from. It's just an accounting method for crediting that an ad impression was made, not for valuing it.
To Joel's, John Grono's, Ed Papazian's, Nick Schiavone's Dorothy Higgins', and even Tony Javis' points, valuing cross-platform video ad impressions will come from other higher-order methods that the MRC and others are in the process of developing now. The duration-weighted impression is just a mechanism for crediting that an impression can be counted, not how much it counts.
If the rest of the process goes the way George Ivie has described it to me, it's most likely that suppliers and demanders will be able to pick their own poison for assigning value to those credits, but the next phase will provide a framework for doing that.
First, thank you, Joe Mandese and MediaPost, for throwing open the windows, once again,
for some fresh air on what had the risk of becoming stale:
the Duration-Weighted Impressions controversy.
The doctrine of stare decisis does not hold in the matter of digital, viewable impressions.
If this case were to be litigated further, as it ought to be in my opinion, there are no media research or financial accounting precedents upon which to base a methodological decision.
I believe, as perhaps Tony Jarvis may, that the closest we come to having core principles
is the original ARF monograph "Toward Better Media Comparisons" (1961).
Twenty-first century attempts to accommodate digital innovation too often constitute "corruption," as opposed to "evolution."
It is with a sense of irony I note that the June 22nd edition of the ARF's "News You Can Use" points to a Member feature titled "Confusion About Duration Weighting Proposal."
Seems like we are NOT dealing with "settled law" (Stare decisis et non quieta movere).
Despite everyone's hope that this topic, like Summer mosquitoes, stays away,
seems like the season has only started.