Commentary
Why Declining Ratings Don't Matter (To The Networks)
- by Steve Sternberg , Featured Columnist, June 1, 2016
If you ask any broadcast network executive whether they would rather lose 10% of their audience next season but move into first place (because everyone else declined more) or gain 10% but slip into third or fourth place, they would all rather lose viewers and be in first place.
Likewise, if you asked every cable network currently outside the top 10 if they would rather gain viewers and stay where they currently rank or lose viewers but move into the top 10, they would all choose to lose viewers.
This, despite just two-tenths of a rating point separating the number 10 and number 40 ranked cable networks among adults 18-49.
This is actually as ridiculous as it sounds, but that’s what happens in an industry that doesn’t reward gains or penalize losses. This is the only business I can think of where losing customers is perfectly fine, as long as your competitors lose more customers.
At the recent upfront presentations, we heard a lot about who is No. 1 in this or that, but we heard virtually nothing about how many people are actually watching all those No. 1 networks.
advertisement
advertisement
We also heard that people of all ages are still watching much more traditional TV than digital or social media. But again, nothing about how many actual viewers that represents.
I’ve been saying for years that this short-term strategy of only thinking about where you rank, rather than whether you gain or lose viewers, could lead to long-term problems. Welcome to the long term.
Ad-supported cable has long been siphoning off broadcast network viewers, but until recently their combined audience remained remarkably stable. Viewers and money tended to shift around the TV pie, but the size of the pie remained roughly the same.
The audience going from broadcast to cable each season was so splintered that no single cable network got too close to any of the Big Four broadcast networks. So the broadcast networks were able to claim that they still far out-rated and outreached cable, while the cable networks were able to claim that in aggregate, ad-supported cable was higher rated than broadcast.
Separate broadcast and cable upfronts create artificial supply and demand enabling the broadcast networks to get higher CPMs from advertisers for declining ratings at the same time that half a billion dollars shifts from broadcast to cable every year.
Broadcast ratings continue to slip, and now non-ad-supported TV/video is finally siphoning viewers from both broadcast and cable.
Ten years ago, I was at a broadcast network upfront presentation where they showed a giant 4.0 on the screen – an average adult 18-49 rating they seemed very proud of, since it was higher than any other broadcast network.
At the time, I wrote in a report to clients that if we continued on this road, they would soon be talking about how wonderful they were to have an average 2.0 rating. I was wrong. They no longer even mention their actual ratings, only their rankings.
Broadcast networks still stubbornly refuse to promote one another as cable networks have done so effectively for years. In today’s media world, is there really any broadcast network executive who believes his only competition is one of three or four other broadcast networks?
Until this changes, the largest group of potential viewers in which to promote broadcast network programming will remain untapped, and network ratings will continue to drop.
And, until the marketplace dynamics and industry perceptions that continue to make audience rankings more significant than audience size change, we'll continue to analyze rankings of continually declining audience bases.
The real question, Steve, is can the broadcast TV networks find a way to significantly increase their average minute primetime 18-49 ratings even if they somehow create "better" shows? It seems to me that this is probably impossible with so many channels to choose from. You mention the decline in their ratings from 4% to around 2-2.5% but over the same ten-year period, the number of channels available per TV home has doubled---which is the main reason for the decline.
I am a frequent critic of the broadcast networks' primetime programming strategies---as noted in my new book, "TV Now and Then"----however, I think that they are never going to see really high primetime ratings again no matter what they try. So, being in first place, which means they can charge higher CPMs than rivals, is not as silly a goal as it may seem.
By the way, there is one way that the broadcast networks and the major cable channels could recoup their lost average minute primetime viewers and then some. If "unbundling" ever comer to pass, so consumers can pick exactly what channels they think they want and pay only for them, the result would be a dramatic reduction in the number of competing channels per home and that would generate a reverse fragmentation effect, causing network and major cable channel ratings to rise---even with the same sort of pap they now offer. I wonder if this possibility has even occurred to them.
ED-
I think the issue with any unbundling startegy is the fact that, on a regular basis, TV viewers only watch 7 to 12 of TV networks available. They already view a restricted network list, so enabling them to only pay for those 7 to 12 will most likely not have the significant boost in impressions they would hope for.
Long
Long, I will shortly be sending subscribers a special ananysis of this subject as part of my new "Mediology" series of single subject white papers. In a nutshell, what will probably happen is that most consumers will select many of the broadcast networks as well as the large cable services---ESPN, CNN, Fox News, Weather Channel, Discovery, TNT, etc. leaving only a few spots to fill for the far more numerous selective channels. As an inevitable result, the ratings for the most commonly chosen channels will rise significantly, because the average home---under the unbundled 17-channel per home theory---will have many, many fewer competing program sources than it has today. There is a clear correlation between the increase in channels and the decline in the broadcast networks' ratings. Under bundling, the same thing will happen in reverse. Fewer channels, means higher ratings for those channels that survive the cut.
What too many people don't realize, is that viewers don't watch or think in terms of networks. No one watches 17 networks, they watch 17 programs. They have their favorites, and when they are not on, revert to their second or third options. That's how off-network series on TBS, USA, and ION do so well - by being the best option when nothing else is on. But when given a choice of channels, people may not think of what they watch when nothing else is on. Folks are in for a big surprise as to how much TV viewing will take place in a 17--channel unbundled home. Higher average ratings per channel, but much less overall TV viewing. If my favorite show isn't on, I'll zip through 17 channels in 30 seconds and then I'm gone. With 4-500 channels, I'll be there much longer and probably find something to watch.
Steve, I agree with you that the availability of more content stimulates greater use of TV and this has been one of the main reasons for heightened TV viewing time, along with the advent of more TV sets per home, which allowed individual family members to watch whatever they wanted on their own "private" sets, more digital options, etc.
The whole idea that people will pick channels, not specific shows---or genres of shows ( pro sports or classic movies, as examples ) is founded on the muddled "channel viewing" concept that is still bandied about by some theorists who need to find thenselves day jobs. Still, imagine that we actually pick only 17 channels to buy as a nationwide average under unbundling and that the big boys--ABC, CBS, NBC, Fox, Turner, Discovery, etc--dominate the choices, being represented on most lists. With fewer competing channels---even if there is a decline in overall viewing tonnage---their ratings will, without a doubt, rise to something approaching their level about 10-15 years ago.
Steve is absolutely correct, no one watches networks, they watch programs. The elephant in the room isn't that ratings that are sliding, it's the outdated metholodology used to collect the data and sample size. Nielsen's samples are small. On a local level, you are looking at 600-800 houshold meters even in major markets to measure households of 1.5 million+. It's hard to call that statistically viable. It gets worse when you look at available ratings points, they are down double digits percentage wise. Are less people watching less or does the lack of mobile, lack of out of home viewing and small samples not giving a true picture of what people are watching. Networks are seeing lower prime ratings and local affiliates are seeing lower late news ratings because of lower lead-in ratings. The only happy group are agencies. They get the benefits of lower costs per point and the volitivity of a small sample provides for bonus spots through under delivery.
Al, are you saying that spot sellers give bonus announcements or "make goods" only when a show undelivers in the ratings but take no credit for over delivery. That's not the way the networks sell. As for Nielsen's small local market samples, no argument there. But Nielsen's answer--- and it's a fair one--- is that nobody is willing to pay the added costs for much larger samples. Is Nielsen expected to lose money by taking that burden on itself?
Ed, I'll take your word on Network selling. UD is not the only reasdon make goods are issued. Bonus wieght is given for a multitude of reasons. But on a local level as far as UD, it all depends on how clients post and their flights. Many clients do not let you cume points for an overall +/-. If you underdeliver, you owe, if you overdeliver, thank you. Now you may get credit as in "you did a good job" and they remember it when the next negotiation, but you may not get it a credit on points owned. As far as Nielsen goes. Aren't they in the business of providing an accurate representation of viewing audience? If they know that the small samples lead to inconsistant delivery and ratings volitivity, then they shouldn't prop themselves up as the perfect solution. They collect millions from stations and produce a product that is often statistically unreliable. Nielsen has no problem saying that X number of millions of people watched the Super Bowl. But when a prime program delivers a hashmark, they say its not that no one watched the program, no one in the sample watched the program. They can't have their cake an eat it too. It's more than just a TV measurment problem. Full disclosure - I am prone to reductio ad absurdum, but here me out. Brand X develops a new product pointed toward young females. They do target testing in markets and buy flights in programs that do well with young females. Product does not sell well and is dropped. Did the product stink, or did it not succeed becuase the advertising was not seen by the target audience due to statistical inconsistancies of a small sample? Maybe Nielsen should think about emilinating overnights all together? When you look at the margin of error on an overnight, it is truly scary. If end clients knew what those margins of errors actually were, I think there would be much less confidence in the numbers Nielsen produces.
Al, I have long felt that Nielsen continues in the local market TV rating business due to its long established standing as the medium's pre-eminant national audience counter. On a national basis, Nielsen's audience size is adequate to report on average minute ratings for most shows---some very small cable channels and a few "long tail networks" excepted. Also, much greater care is devoted to keeping the national panel as up to date as possible---currently expanding to include reporting on digital venues plus some out-of-home locations. But local TV lags well behind because---let's be frank--- that's where it ranks with the major ad agencies. Meanwhile, the stations---for understandable reasons, won't pay more for their ratings.
I believe that Nielsen should seriously consider working out a deal with comScore, which now has Rentrak under its banner, to deveolp a large sample local market service which utilizes Nielsen based viewer-per home data in conjunction with comScore's set usage and digital findings to service the rating needs of spot sellers and buyers. In fact, I wouldn't be surprised if both companies eventually merge into one entity. There are many technical issues to be resolved but I believe that it can be done. In the meantime, we are stuck with a very poor local market TV rating service and, believe me, I feel your pain.
Al, regarding audience guarantees, the TV networks often sell program packages involving many shows and telecast dates and they guarantee GRPs not show by show or by telecast but for the entire combined buy. That's how they reap the bebefits of over delivery as well as accounting for undelivery. I understand that spot handles this differently and, in my opinion, unfairly to the sellers. Isn't there a way to adopt the national selling practice whenever time is bought in more than one show?
Ed, a merger between Comscore / Rentrak and Nielsen would be a huge improvement. They maybe able balance out each others methodology warts. Nielsen could be pushed into better local service because of the Charter TWC/BH merger. Rentrak had a contract with Charter (one that will sure be renegotiated because the value of Charter's data just got exponentially better). If say Rentrak can deliver a sample size of 1/2 of an individual market's cable and ads subscribers, does it become the more valuable measurement - even with it's modeled demos?
Al, regarding my hypothetical Nielsen-comScore local market TV rating service, clearly a much larger set usage panel would be an improvement---providing adjustments are made to refelct the fact that non-set-top-box homes need to be represented in the data and how other more technical issues about panel turnover, how average commercial minute ratings are tabulated, various weighting schemes, etc. are dealt with. The key would be how the viewer-per-home data was integrated with the set usage findings. Even though it is based on inflationary assumptions---Nielsen doesn't really know whether its panel members actually watch commercials--- I would hope that this would be done in exactly the same way that Nielsen's national ratings are tallied, for comparability purposes.
One final comment re sample sizes. Larger samples, per se, do not automatically produce more accurate data. Other factors, mentioned above, as well as how the samples are recruited, whether they are representative, whether all set usage is measured, etc. come into play. Also, the viewer-per-home part of the equation would be based on much smaller samples than the set usage component. Still, I believe that a combination of the two methods, while not the perfect solution, enables station time sellers an opportunity to greatly upgrade their rating studies---hoipefully, without the huge cost increases reqiuired to fund Nielsen-style peoplemeter panels of adequate size.