But much the way the TV universe has evolved in terms of the types of programs -- and platforms -- that critics review and people watch, it has also been affecting the statistical way ratings are actually calculated, and much to the chagrin of programmers, those changes usually mean increasingly lower ratings.
Part of that trend is related to a real programming phenomenon: the golden age of Peak TV has created so much watchable programming that audiences -- and ratings -- have increasingly fragmented.
But part of it has to do with methodological changes in the way companies like Nielsen estimate the ratings of television programs.
Remember, a rating point represents a percentage of viewers available in the television universe -- whether they are actually watching something or not.
And in recent years, as more and more viewers began watching TV in non-traditional ways -- namely over-the-top, via the internet, on connected TVs, etc. -- Nielsen has had to effectively redefine and recalibrate the way it defines the television universe.
The last big change was when it began including broadband-only households -- homes that have no conventional linear TV access, but may or may not watch TV online or via connected devices.
TV programmers resisted that change, because it increased the size of the denominator that ratings are derived from, making them look even smaller.
Today, the Advertising Research Foundation (ARF) is releasing new data -- as well as a call to the advertising and TV industry -- to change the definition of the TV universe once again: moving from the concept of "TV households" to "TV-accessible households." Or in the parlance of ubiquitous industry algorithms, "TVAs."
Based on the ARF's new research estimates, 5% of U.S. households currently do not even own a TV set, but 4% have broadband access and therefore theoretically are TV-accessible under this new definition.
The new definition makes sense, and is more representative of today's actual "TV universe."
But the ARF's call nonetheless will be controversial among TV programmers, because it will also likely have the effect of increasing the denominator ratings are derived from and reducing the ratings reported for any given television program.
So here's my solution: Maybe it's time to scrap ratings altogether, and just use the number of people who watched a TV show?
The truth is the ad industry has already moved away from estimating campaigns based on so-called GRPs (gross rating points) and simply calculating the number of impressions (viewers) reached.
So maybe it's time for the rest of us, TV critics included, to say RIP to TV ratings.
More on programmers stating currency preferences, the role ACR data is playing in CTV measurement and why the definition of currency is not neutral but a point of contention between programmers and digital video natives like YouTube.
Unfortunately some of your references are not sound. As Ed Papazian, John Grono and others have commented many times in Media Post, current US TV ratings do not measure "actually watched" or "actually viewed". Or, as we say, do not measure Eyes-On/Ears-On or contacts which is the vital pre-attentive processing measure. Note: The currency for OOH has been at the Eyes-On level globally for the last 14+ years!
Based on that understanding, your "Just use the number of people who watched a TV show", is on point as long as it is not twisted and falsely based on ACR which is only a device/screen based metric with no people "watching" or Eyes-On measure whatsoever, reflecting Benny's point.
This ARF recommendation adds another dimension to the currency wars which should be striving for neutrality but must ultimately serve programmers and advertisers with menaaingful real consumption data. As such, only measurement of actual persons with Eyes-On/Ears-On or contacts at a minimum, or better "Attention" or impacts should be the basis of any medium's SINGLE currency. As ARF includes research vendors as members, one would hope they collaborated closely with MRC which does not allow research vendors to be members and always strives for neutrality in accredited measurement approaches and resulting metrics.
Joe, just because the number of households in the projection base is expanded by a small amount that is no reason to cease expressing "audience" in both ways---the percentage of homes or viewers who might have "watched" and the actual number of same who the survery claims "watched". It's the same issue I have with the great leap sideways with some folks thinking that GRPs are outmoded and "impressions" are a better way to go. The average media or ad executive does does not think in millions ----or billions--- they think about scale---whether it is rech or frequency or share of market---- in percentages as these provide a common and readily understandable frame of referrence not only for current deliberations but for comparisons with the past. Hit them with boxcar stats and all they will notice is that one stat is bigger or smaller than the other---but they will not be able to use such information as readily as percentages in their thinking process.
There is an increasingly worrying addiction to "the biggest number" in media measurement.
In today's and recent MediaPost articles, there is a cornucopia of opinions of how to measure media usage. Is it impressions? Do we dump TV ratings? Let's just use detailed data sources? Let's also not investigate their veracity. Let's also not standardise the time period over which we are reporting ... you'll get a bigger number if you report 'per week' or 'per month'.
There seems to be a misunderstanding between 'accuracy' and 'representative'. For example, the ARF's call is to scrap TV Households and replace them with 'TV-Accessible" ... to gain what appears to be 4%. What a windfall to get the bigger number! I agree that NOT including 'TV-Accessible" reduces the reported audience. But let's also consider viewers who were watching/listening but not in their home, but at a friend's abode, a club, a pub etc. So let's go chase that audience in the quest for the holy grail of the biggest number.
The problem is that the cost for those relatively small increments in the quest for the biggest number are disproportionately expensive in many orders of magnitude. But sure, if you can afford it, we can do it. And you know what ... the financial gain will be small (if any at all).
Similarly, we're comparing TV ratings average audience data (i.e. if you only watch half of a program you only count as half a person) with digital data where you only need to have viewed/used for a second or two (don't blink - you will be counted) and you are considered to be of equal importance as someone who viewed/used the entire content.
So I think it is more urgent at this point in time that we focus on the data sources being equally represented, and once achieving that we tackle the accuracy issue.
Agreed. It's time.
John, what is, in a way, amusing, is the sellers' nearly pathelogical fear of smaller numbers---and, be assurd that they are the primary force behind the drive to get larger and larger reported "audiences". The sellers don't seem to get that your typical branding advertiser belives----and not without good reason----that the "TV" manner of communication is the one which will get the best results. So these folks are wedded to "TV". And fragmented ratings, aside,the inflated "audience" stats they they see in their media plans still give them comfort that they are spending enough---and getting enough value for their "TV" ad dollars. But if they knew what the truth was---about how little their ad reach actually is and how rarely their ad messages are seen, contrary to what the sellers think, most branding advertisers would seriously consider spending more on "TV", rather than less. They have nowhere else to go.
I agree Ed. But, I have a 'but'.
You commented about inflated data and if they only knew about "how little their ad reach actually is".
Being parsimonious, 10% reach for a TV campaign is probably considered small by most marketers. But in a 312m TV population we're still talking about 31m people!
That is where people come undone when making comparisons and often go with 'the big figure' and miss the opportunity with 'the big figure. It is even more important in global analysis. The 10% above would be 140m if it were China.
Luckily, John, it's not as bad as your hypothetical example of 10% would suggest. When we look at the reach versus frequencies in today's "TV" media plans and factor in attentiveness to commercials the biggest loss is in thefrequency, not so much the reach. For example a 60% monthly reach and a 4.0 averge frequency tallied the old way---supposedly OTS---though not really--- probably gets you a real reach of 40% and a real average frequency of 2.1. So your loss in reach is about a third while your frequency drops almost 50%.
I think that the demos should be done with when it comes to TV ratings which shouldn't go away as TV ratings still matter. I don't get why 4% don't have TVs as I like to watch shows & sports on the big screen and I can find almost anything on TV.
I've been waiting to see if anyone mentioned that Nielsen already includes these homes in the denominator. Has done so for years. and it counts viewing in these homes too. So what are we talking about here? Other ratings/audience measurement providers? Did the ARF forget? Did they check?
Jack, as a sidebar on this subject, when Nielsen announced that it was planning to add broadband-only homes to its local market samples---as it has with its national panel---the stations rose up in arms and violently protested against this daring move. Why? Because their ratings might drop a tad. Yet it's been the stations who forced Nielsen to cease reporting ratings locally and go to "impressions"---which they seem to think is something new and more "comparabe" to digital "audience" reporting. Didn't they realize that adding broadband - only homes to their local survey bases would result in slightly more "impressions"?
Thank you Jack for highlighting that those with the squeakiest wheels should pause, read the manual, and then put some air into their tyres before they complain.
And yes Ed you have reinforced my concern that the race to get the biggest number is devaluing media measurement per se as we sink to using the lowest common denominator, thus devaluing not only themselves but the whole industry.
I do recall that Ed, and also wondered, having started in the stations tv business years ago. The answer was that local market currency is based on cost per point (CPP) and not CPM and the change could be confusing to mom and pop clients. Of course this is just another way to try to maintain the status quo and pretend the world won't change. Even the smallest change will be blown out of proportion to those with vested interests, and the power to obstruct change.