There is no greater inertial force in the world’s nearly trillion dollars of annual media and marketing expenditures than the resilience of the sex/age demo-defined currency that drives virtually all U.S. television ad spend.
Even folks in the industry with their proverbial heads in the sand -- or maybe especially those folks -- know about the recent controversy about Nielsen's panel-based national television advertising currency and its suspension by the Media Rating Council due to data collection issues during COVID-19 lockdown. This controversy will neither kill the Nielsen panel nor its importance to today’s TV industry.
However, if we are lucky, the current kerfuffle between Nielsen, the Video Advertising Bureau (VAB) and the MRC will kill the real nemesis to modernizing the TV and video ad industry -- the sex/age, demo-defined gross rating point (GRP) as the primary planning, buying and measurement metric. Instead, long live the digital and data-defined audience impression as the primary metric for the future of scaled, premium video advertising on TV and similar high engagement video advertising.
For decades, marketers and media professionals have bemoaned the fact that they needed to convert tightly defined strategic marketing targets - affluent, young car buyers with a desire for sports cars, for example - into blunt, broad, fluffy sex/age demographics - Adults18-34, for example - as the primary metric for the planning, buying and measurement of their TV ad campaigns.
Sure, agencies and networks could apply filters on top of those to try to better ensure that they could maximize the delivery of those strategic targets; but good luck getting the networks or agencies en masse outside of a couple of advanced TV ad players to guarantee the campaign against the tight strategic target. And even tougher luck if you want the kind of reach and frequency guarantees that you’re used to getting from pure digital ad players.
That world is now going to change -- and change fast, I hope. Nielsen is already working on its replacement to its legacy TV ad panel called NielsenOne with the promise of granular, cross-screen metrics for planning, buying and measurement. The VAB and its member TV networks are all now promoting audience-based impression buying as their future. And the MRC is already certifying digital ad platforms and suppliers for cross-channel granular audience measurement.
I’m not Nielsen’s apologist, but the enemy here isn’t Nielsen. They are the messenger. It has been the media industry’s decades of resistance in leaving the sex/age demo GRP behind as the core metric. We’ve needed to rip the band-aid off for years. Finally, everyone is in agreement to do it, and to do it now. It will be a nice thing when it happens, even if it creates some new winners and losers, but that’s a column for the future!
Dave, even though many people are under the impression that 18-49 or 25-54 re a targeting metrics, they aren't. these are merely a compromise between seller and buyer to create a single, broad based "demo" to be used for audience tonnage guarantees for the buyer's protection which were developed long ago when the TV networks had many shows that appealed to younger viewers as well those for older viewers. That's all. If you want to have such guarantees you need a single metric as more than one will not work---which means that currently, the best way for most TV sellers to go is to switch to an even broader "currency"---like adults 18+ or even persons 2+. That, at least would account for all of the viewers---or most of them--- not a small segment which is what 18-49 or 25-54 now represent.
As for GRPs, these are not going away as thye are the easiest way for agency clients to develop reach goals and calculate frequency tonnage---or the mythical average frequency---- of their planned schedules. If an advertiser was able to define TV show audiences by product usage or some other characteristic---say audiences who are fashion conscious or price conscious the media plan would still express its reach/frequency goals in percentages with GRPs as the base and buyers would be trying to buy time to attain those GRPs ----not "impressions". Whether the sellers would offer audience guarantees on such customized metrics is the key question---especially for upfront buys where many of the shows involved haven't even been produced as yet. So in addition to not balming Nielsen, we should also not blame GRPs---they are merely a convenient and more understandable way to express audience tonnage---for any targeting metric.
AMEN! You know better than most, but there is clear evidence that advertisers are leaving TV ROI untapped by using antiquated age/sex demographics.
Howard, all of us---You, Dave and myself as well as lots of others agree about sex/age time buying---it's next to useless for targeting purposes. But does it matter? I maintain that most national TV advertisers believe that the main point is to get lots of eyeballs at the lowest possible cost and their commercials will do all of the work from there. However, and this is the key point,even if one found a way to define TV aduiences, including predictions for shows that do not even exist yet in upfront buys---on a more selective and ad-relevant basis---and this may be possible---you still have the issue of the sellers controlling the packaging and the pricing. So, if even if they allow you to cherry pick among their shows, the likliehood will be that the same programs will index high for many advertisers, hence the sellers will instantly note this surge in demand---and react. How? By dramatically raising their CPM prices for spots in such shows. Result: your scientific, modern, targeting approach, which we all agree is better---in theory----yields you fewer targeted ad exposures than buying time tho old fashioned---and, not so smart---- way. Which approach has the better ROI? I think that the issue is in doubt.
At the end of the day, GRPs and impressions aren't very different from one another, are they? Neither is any good at disentangling frequency from reach.
Thanks so much for the thoughtful commentary. As you point out, sex/age demo metrics were never intended as targeting metrics, but they did become that de facto. I do agree that the devil is in the packaging of campaigns and believe that as networks open all spots to a more granular sale, we will see more dynamic pricing mechanisms that will be based on a brand basis that will yield better results for both advertisers and networks. Alas, that will work against the corporate entity pricing, but I am hopeful that we can get there some day.
Kill the GRP? Heinous suggestion! The issue is not the GRP, the issue is the insufficient financial support to properly measure the rating point in a fragmented TV ecosystem.
Suggesting that we throw away the GRP endangers the notion of disregarding effective reach, frequency management, and cross platform duplication and deduplication. Careful not to throw out the baby with the bath water, Dave.
Thank you Mark. As you know, i totally agree on the notion of the power of total points, reach, frequency and audience targets. But, as you point out, as marketers started caring less - and paying less - about how well those GRP's were shaped into actual campaign level reach and frequency, it become only about the GRP. I am fine with the GRP living. I just want to take its proper place as "a" metric, not the only one :-)
Setting aside my own bias on the topic it's delightful to see such smart, open debate on an issue as pressing as this one. Wasn't long ago when there was no debate...nor any viable alternative(s) for measuring TV. In examining the issue I would only suggest that we bifurcate "measurement" from "currency." Regardless of the basis by which TV spots are bought and sold there's absolutely no reason NOT to use more granular (impressions-based) metrics to determine which spots to buy or sell....and/or to calculate what the effective rate (eCPM) is for reaching luxury auto-intenders on broadcast prime vs daytime vs OTT. Even if it takes time to transition to a more unified impressions-based currency that doesn't mean we can't change the basis by which we measure performance and outcomes.
Focusing on low cost impressions is the easiest and least expensive to do. It is easy to understand and execute. Focusing on ROI leaves money on the table because you can skim the cream but leave lots of money on the table. Neither maximizes profits. The closest we've had to that is cost per incremental reach point.
Jerome, in my mind, the most important difference between GRP's and impressions is granulairty. The impression is a sub-atomic-unit of the GRP. Starting with impressions enables you to build plans with people first, not just programs, netwokrs and dayparts. Starting with GRP's, that are not divisible, means that you are stuck with lots of averages. Thus, most larger TV ac campaigns today deliver a massive disparity in frequency by audience member - typically, with the heaviest hit quintile getting 25-50x the frequency of the least hit. Srating and ending with impression metrics exposes this. Working on GRP's as the core model enables folks to look at the "4x" average frequency of a campaign like that and totally ignore the 25x disparity.
Demographics have always been proxies for targeted audiences. Ideally, if I'm selling cheese, I don't care if you are 6 or 60; if you like cheese and have money in your pocket, I want to talk to you about my cheese.
The GRP is a kind of compromise between buyers and sellers, as Mr. Papazian points out. And it does make some sense, in light of what most advertisers want to accomplish (mass reach for acquisition), to interogate whether or not a more precise form of currency is warranted. For most mass produced, mass marketed product, the compromise that's been worked out between media owners and media buyers works well enough. If you are McDonalds, you really just want to reach anyone with a mouth. If you are Colgate, you want anyone with teeth in that mouths. If you are Charmin... well, you get the picture.
I do think that as more devices become connected, thereby rendering media vehicles more data-rich, a move to audience-based currency can happen. It might not kill off the rating point, but it might make TRPs of greater targeting texture easier to come by. In order for a system based on that level of discernment, common ground on how the data is collected and rendered has to be found, but I don't think it's beyond the pale for current systems to do that already. Nielsen's moving in that direction themselves; they just want to stymie the rest of the ecosystem from coming to that common ground without them before Nielsen itself can colonize it.
The whole debate about the usefulness of GRPs continues to surprise me. A rating point is not a consumer---you can't target it or send it a commercial or expect it to respond by purchasing your product. It's merely a statistical way of expressing a unit of audience---in this case one percent of some base population or target group. As such it may represent a 1,000 people ---or 10,000, or 100,000---depending on what the base happens to be but these are not specific people either---it's a tally of audience" impressions". That's all.
Most CMOs and brand managers think of media plans as generating a certain level of reach in a week or, more commonly, a longer period like a month. Unlike direct response or search advertisers, branding advertisers do not specify the amount of "impressions"---supposed audience contacts---in millions or billions ---nor do they evaluate their campaigjns based on click throughs or "conversions". Their advertising is not specific to each individual---though the dreamers hope that one day this will be the norm. Rather, they are targeting types of consumers---sometimes several or overlapping types. Having defined these types as targets and quantified them in aggregate as representing a certain number of people---many millions, usually---they plan to attain a certain coverage level---or reach---for their ad campaigns.. This is expressed as a percentage. In order to calculate the average frequency---itself a misleading stat---you need audience tonnage expressed in the same way---hence rating points.
No matter how sophisticated we get in terms of targeting and media buying, we will still make plane based on anticipated reach and these will still be discussed as percentages---so if you want to talk about frequency---even if we are deluding ourselves that the audience surveys measure actual ad exposure---you will need GRPs.