Commentary
Three -- Or Four -- Planning & Buying Issues Worth Arguing About
- by Joe Mandese @mp_joemandese, April 18, 2024
That's what my colleague Steve Smith just asked me as we begin finalizing an agenda for our first-ever Planning & Buying Insider Summit in Austin later this year.
Full disclosure: I'm not an industry insider, I'm just a journalist who writes about them. So I'm doing what journalists do and asking the best sources I have -- you.
Based on some recent conversations, pitches and observations, I'm using today's edition to tease a few subjects out, but what I'd really like is for some of you genuine insiders to give me feedback, shoot them down, refine them, and/or suggest some topics that would be -- in Smith's words -- "fruitful" for debate.
As always, you can post them as comments below, or email me directly at joe@mediapost.com.
advertisement
advertisement
For guidance, Smith says "We are going to go meta and micro," with keynotes and panels tackling broad strategy questions such as "allocation, measurement, AI, creator
economy, etc. and then we want cases that drill into emerging media categories like metaverse/gaming/DOOH/Audio/
Now here are the big planning and buying questions on my trade reporter mind. Spoiler alert: I've already written extensively about some of them.
- The universe. No, not the astronomical one, but the media one. It's about to change in a fundamental way. For most of my career covering the ad business, it was framed around Nielsen's TV universe, which so far has been the base buy for most national advertising plans, even though there was an undefined digital universe expanding -- and some might say, overtaking -- it. Based on Media Rating Council briefing I sat in on late last week, the TV universe will effectively disappear with the transition to "cross-platform" measurement beginning this year. And with it, the MRC has recommended the ad industry move from siloed media ways of thinking about universes, to one based on the general population. You know, the Census. That makes sense for a lot of reasons, but what I'd really like to understand from all of you, is how does that shift impact the way you think about, plan and value media? If TV really has been the lead dog -- the base buy -- framing the way national media plans are conceived, what happens when it's just a part of the mix. Is this finally the end of a TV-centric planning mindset? Will it be replaced by a cross-platform one, which seems to be the goal of OpenAP, the U.S. JIC, and the big TV/streaming-centric suppliers driving it? Or will a truly agnostic and holistic way of planning media grow out of it. You know, one based on a consumer-centric way of thinking about media? I could go on, but I want to save some of my questions for a fruitful debate on stage in Austin.
- The "dark universe." Personally, this is one of my favorite theories about the media universe -- that for all the metrics and methods the ad industry uses to define media, there are vast pockets of unmeasured media influencing both consumers and brands -- but they don't get on the plan, much less the buy list, because they are not measurable, and therefore, not buyable. At least not in the way big advertisers and agencies are set up to plan, buy and post the impact of media. Needless to say, some forms of the dark media universe have come to light in recent years vis a vis new ways of reaching and influencing consumers on behalf of brands. Things like "native content," influencer marketing, the "creator economy," direct-to-consumer marketing, "retail media" (see separate bullet point below), but we're still only scratching the surface according to a state surfaced by Madison and Wall analyst Brian Wieser, who while still at GroupM, came up with an estimate that marketers spend $6 trillion a year on marketing overall. That's at least twice as much as what most industry bean-counters have previously benchmarked, which means there's an awful lot of media not being accounted for in most advertisers, agency and trade publisher discussions. To be fair, much of what Wieser says is in those trillions of unaccounted for marketing budgets, probably isn't even measurable (think sandwich boards, flyers and forms of direct selling and unstructured media that nonetheless are part of the total marketing universe and in some ways or another compete for both budget, as well as consumer attention, engagement and fulfillment.
- Retail media. One of the most fascinating "new" categories within the media universe is the relatively recent benchmarking of a humongous "retail media" category. By some estimates, it's already more than $100 billion and at least one big agency (GroupM) forecasts it will be close to half a trillion dollars and will overtake TV ad spending soon. The funny thing, is it's not actually entirely new. It's just emerged in a new form -- digitally via ecommerce sites, both pure-play, as well as digital extensions of physical retailers. Truth is, it's not entirely new. When I first began covering media in the early 1980s, most marketing spending beancounters understood that "below-the-line" marketing already was twice as big as the above-the-line one: ad-supported media. Within the below-the-line universe was things like PR, direct marketing, and consumer and trade promotion, which all have taken on new forms in the digital universe, including retail media, but they're just new names -- and often new players -- doing what marketing has always done. My question, is how much is the newly defined category of "retail media" competing with, or incremental to the rest of the ad-supported media marketplace?
- KPIs, outcomes, etc., etc., etc. Perhaps the most fundamental shift I've observed in the way media is planned, bought, sold and measured is the business objectives used to define the ROAS (return on ad spending). The simple description is that we've shifted from long-term brand-building attributes like brand lift, awareness, recall, consideration, affinity, etc. to short-term ones like conversations, sales, opt-ins, etc. The truth is, it was never binary and both long- and short-term outcomes are important for brand marketers, but as the ad-supported media industry increasingly shifts to "outcome-based" metrics and measures, I for one sure hope they don't throw the baby out with the bath water.
- Contextual vs. behavioral. Needless to say, this relates to the outcomes debate as well, but there's a separate debate about whether it's better to target consumers based on their personal identity data and behaviors, or on the media they are using. While the deprecation of cookies and device identity trackers has accelerated this debate and contextual recently has experienced a resurgence because of it, it seems to me that the industry remains committed to identity/behavioral-based targeting and that it's just gone from a passive means of acquiring it (dropping browser cookies and devices IDs) to an active opt-in model managed by brands, agencies, platforms and media companies all claiming to have the best, most privacy-compliant and ethical identity spines. Not to mention a burgeoning cottage industry of data cleanrooms who can mix and match all those sources of data to achieve optimum "identity resolution." Is this really what consumer advocates and regulators intended, how long will the identity spine marketplace sustain itself until there is a new spate of regulation? And what will the role of contextual targeting be within it?
- Media decentral. The sideshow of decentralized Web3 models, the blockchain, and consumer data sovereignty may seem like some unneeded noise for planners and buyers, but to those leading the charge, it promises to be every big as transformational as HTML was to the formation of the current digital media marketplace. Sure there's a lot of hype and speculation -- particularly among crypto hedgers -- but there are some fundamentally new ways of thinking about how consumers and brands will interact within some of these new models. This one may be too hot for a fruitful Planning & Buying Insider Summit debate, and in my humble opinion, is deserving of its own, separate show. But at the very least it would be interesting to see how some traditional media thinkers think about the implications for a decentralized media marketplace controlled by individual users, or new custodian platforms, as well as some likely Trojan horses that emerge along the way.
- Information asymmetry and misaligned incentives. I probably sound like a broken record by now, but truly believe these are the two most important ad industry issues identified by the Association of National Advertisers in last year's transparency reports. And for what it's worth, I don't see how the sell-side and agency-side push to create multiple new advertising currencies helps solve that problem, and I believe will only exacerbate it.
- The power of Babble (see above).
7 comments about "Three -- Or Four -- Planning & Buying Issues Worth Arguing About".
Read more comments >
Next story loading 
Joe Mandese is the Editor in Chief of MediaPost. You can reach Joe at joe@mediapost.com.
Joe, I am confused by your description of the change in our "universe" definitions for measurement of "audience". When Nielsen set up its radio meter panel in the early 1940s and later, simply convertied it to a TV rating panel as various membhers aqcuired their TV sets, it was designed to represent the entire TV household universe---including, in theory, those with TV sets who almost never watched TV. The same objective was evident in all of the other surveys---Arbitron for TV and radio, Simmons, TGI, MRI, etc. for magazines, etc. Each was trying to attain a nationally representative panel---or sample. It might be argued that this is not attainable and that none of these surveys was able to obtain a perfect national sample---which is true. So they applied sample balancing to account for having too many---or too few of a particular demo---a standard, albeit not a perfect solution.
So what's new?
All that is happening in "TV"---which is linear and streaming, not just linear---is that thanks to audience fragramentation---we are going to much larger panel sizes. But the core idea is still exactly what it always was---trying to obtain a nationally representative panel---or sample. The problem rests with the use of a variety of existing, hence cost efficient panels---none of which are necessarily representative of the universe they claim to measure---usually all ACR or STB homes. So those, like Nielsen, who are trying to meld these together into a nationally representative sample have theuir work cut out for them---and we will have to wait for the MRC to sort all of this out.
I suspect that some of us are confusing sample size with accuracy and, accordingly, when we say that TV audience measurement is shifting to a "census" style approach, we are thinking that this represents a change and that old, small sample surveys were not representative ---or accurate. That simply isn't the case. When Nielsen finally begins to release network by network and show by show results from its new, "big data" plus people meter panel and comparisons are made with the findings of its current "small data" panel, they will probably be substantially the same. If that is so---exceptions for extremely tiny audience, "long tail", programmers, notwithstanding-----what are we changing regarding our "universe"?Nothing. We are still trying for perfection---truly representative national samples---but probably not getting that---just as in times past..
Joe, regarding the shift to "outcomes", I would note that there are basically two main types of advertising---although these sometimes overlap. One is what might be called pure branding---of the kind most commonly seen on "TV" ---but in other media as well. The other is more sales promotional in nature and is represented by direct marketing and its close relative, search. The last one---search--- which accounts for a huge share of digital media ad spending, is primarily based on "outcomes" as most buyers pay only for what they believe to be valid clickthroughs---though not usually for "conversions". So that represents a change though it really is a substitute for much of direct mail's ad dollars which were also based on response metrics.
I think that many are under the impression that the national TV time seller orchestrated "JIC"has set in motion a move to "outcomes" for TV but this is a very dubious notion. Why? Because the various add-on metrics "certified": by the "JIC" are purely optional, not standard options. Also and not surprisingly, they are being promoted by those sellers who believe that their sales strategies will be enhanced by them---the buyers don't get to tell each seller what types of "outcomes" the seller should guarantee performance on. As a result, these add-on- to- Nielsen metrics will be employed selectively and the sellers will make guarantees only in the context of what might be realistically expected. So most buyers will get just about what they would have gotten anyway---or, maybe a tad better------and probably at a slightly higher CPM to pay for the added research.
The whole question of "outcomes" needs to be considered in terms that most sellers could live with. The digirtal media seller who chunks out huge numbers of supposedly targeted "impressions" gets paid only if there is a minimal response---like clickthroughs for search. That's fine---but will this now shift to somewthing more meaningful for the buyer---and riskier for the seller---like sales results? And the same question might be asked for TV branding campaigns. If a seller guarantees a particular type of "outcome" response rate for some buyers----but not sales results ---is that really a sea change---or, maybe just a small, ripple in the wide ocean?
@Ed Papazian: Historically, TV audience estimates were derived based on the percentage of total TV households that are viewing television during a given period (households using television % = household rating). Ditto for PUTs (persons using television) % = # of persons viewing TV.
The new cross-media measurement standards call for real, individual people projected to the total U.S. population (or relevant geo), not TV or digital populations.
Joe, changing the "base" to total households---or people---rather than those homes that own at least one working TV set should produce almost exactly the same estimate of "homes tuned in" or "viewers". The only difference would be that a very, very small number of people who don't have their own TV set might be reached in an out-of-home location other than the most common OOH viewing situation-- a set in another person's home ( viewing by "visitors" in homes has always been included by Nielsen and Arbitron ). It's also possible that a very small number of homes or people who don't have their own TV set might be reached via a digital device. But how that is dealt with I have no idea. Carried beyond this why not include "non-household" residents in "group quarters"---like prisons, nursing homes, etc.--- which are also excluded from the measurements as being too difficult to manage properly.
Frankly, I am amazed that people think that trying to include every tiny bit of "audience" when we don't have a measurement meethod that tracks second by second"viewing"---in-home or otherwise---is a major improvement. For example, how representative are those seemingly gigantic ACR home or STB panels with their millions of members? What are their opt-in rates? How do the ACR panels account for non- ACR homes or old fashioned sets in their panel homes? And, most important, do we really believe that having a TV set tuned in when a commercial is on-screen automatically means that the program viewer is present and "watching".
Most time buyers---when they think about it realistically--- know that no TV rating system is going to provide absolute truth about the size of the "audience" or its exact demographic makeup. What they are concerned about is how whatever is being measured is divided among the various networks, channels and platforms and within those parameters, among the various programs that carry ads. So long as they believe that the ratings are not artificially tilted to favor one show over another or one seller over another, they feel that they can make buys for their clients that are reasonably objective. In that context, the endless but basically hopeless quest for perfection and 100% completeness is a nice dream but meanwhle life---and time buying---- must go on.
@Ed Papazian: I think you're missing the point, because you are thinking in a TV-centric way.
The point is the industry would be shifting from a TV-centric -- or a digital-centric -- way of thinking about the media universe to one based on the actual population.
Whatever the margins of difference might be vis a vis a medium's universe (including ones yet to come, ie. the metaverse, Web3, who knows what), the point is creating an indivdual persons-centric framework based on the population of people who use media.
That is the goal of the new standards.
Adding this link so you can see the MRC's crossmedia measurement standards framework:
https://s3.amazonaws.com/media.mediapost.com/uploads/CrossMediaMeasurementStandards.png
The point isn't to create a new crossmedia universe, but to shift from a media-centric framework to an individual persons framework.
Joe, I remain mystified---sorry.
If all of our supposed "audience" measurements are telling us how many and what kinds of people are "exposed" to each media vehicle and its ads, why is that different from what you are saying?
I assume that the answer is that we are supposed to somehow shift to knowing exactly what products/brands each individual is buying and the media he/she is exposed to and tailor our media placements accordingly. How that miracle is to be accomplished with the crude measurements now in use is beyond me. Certainly, monitoring the usage of every ad carrying device---if this was even possible----would be a first step. But that would be at best a very misleading indicator of viewing exposure to both content and ads.
This isn't just a "TV" issue; it applies to all media. However, I refer, again, to my earlier point about branding ads versus direct response ads. They march to different drummers where measurement and "outcomes" are concerned. If a search advertiser pays a media seller only when a valid response is attained---though not necessarily a sale---why does the advertiser care so much about exactly who might have seen the ad ? In contrast, a branding advertiser, while welcoming any instant sales response, is paying for each ad exposure and expecting each one to contribute to the desired long term effect of the total campaign. So targeting and issues like attentiveness, how the ad messages are scheduled, etc. become of prime importance. Is it really essential in either case to bore down to the individual consumer level?Maybe the answer is yes for search in the sense that you go back to those who respond with new offers---but does this apply to branding advertising where you are dealing with huge numbers of consumers who might be persuaded?I'm not so sure about that.
@Ed Papazian: We're talking about two different things in this thread. I was only responding to your comments about universe estimates when I referred to your TV centricity, not to outcomes-based measurement. Apologies if that wasn't clear.
But all your comments affirm the goal of my column, which was to see if these are planning & buying issues worth arguing fruitfully about.
I'd love to know what others think.
And I'd love to know if there are other issues you think would be worth arguing about at MediaPost's upcoming summit on the subject.