Why Have TV Viewers Stopped Channel Surfing?

  • by , Featured Contributor, November 16, 2017
We all know the headlines about television these days: “Massive ratings declines.” “Prime time down.” “TV dying as viewers cut the cord.”

Do these really tell the whole story about the behaviors of U.S. TV viewers, and the health of TV as media?

To find out, my company’s chief scientist Kyle Hubert recently teamed up with Forrester’s Jim Nail to dig into the numbers behind the ratings. They analyzed the “gold standard” of TV viewing data – Nielsen’s All Minute Respondent-Level Data – over the past four and a half years to better understand what was happening at the individual viewer level.

 The high-level findings were what you’d expect. Almost everyone watching TV has access to more channels today than they did five years ago. The ratings of top shows have declined over those same five years, on average about 25% down. To reach the same size audience as five years ago advertisers must buy spots on significantly more shows and  networks today.



 However, a lot of what the data revealed was not nearly as aligned with conventional wisdom about what is happening on TV.

First, while individual show ratings are down by 25% over the past five years, viewing per person has only dropped 15% over that time frame. How is that possible? It’s simply because more shows are being viewed by the average viewer than five years ago. And, if you look at TV viewers who are 25 and older and watch at least 12 minutes of TV a day, there is no change in the amount of TV they watch. Yes,  25+ TV viewers are still watching the same amount.

Further, total TV audience viewing is up 5% over that time, since more people watch TV today. And like the U.S. population overall, the TV viewing population has grown.

Also, the number of different networks each viewer is watching has dropped over the past five years. Sounds counterintuitive, but it’s true. Most TV viewers watch only one or two  networks each day. The number of people who only watch one network per day has doubled, while those watching two networks per day have almost doubled. The number of those watching three networks per day are up 50%.

Wow. Shouldn’t more choice mean that viewers will watch more networks? No. What has dropped is channel surfing. What you can see in the respondent minute level data, and is confirmed by investigation of second level set-top-box viewing data, is that fewer viewers are surfing up and down channels. There is less “snacking” on TV today than five years ago.

In fact, most viewers today ave fewer viewing event and watch for longer periods of time.

With viewers now watching fewer networks and watching each network for longer periods of time, is it getting easier for advertisers to reach them? One would think that a TV media buyer today could just buy the top 10 networks and get the most viewers.

Actually, it’s quite the opposite. TV viewers may be watching fewer networks, but they are selecting those networks from a much larger pool. TV viewing has become more individualized and much more fragmented, with apparently more direct navigation to channel and show destinations. To reach all viewers today who only watch two networks (the largest group of TV viewers), a marketer would have to advertise on 135 different networks.

This has enormous implications for the TV media business. Thirty-five years ago, an advertiser buying one spot each on the top three networks in prime time on the same night reached virtually 100% of all TV viewers. Today, that buy yields less than 5% of TV viewers.

These findings turn upside down what many have thought about TV viewing. Anyone who wants to survive and thrive in the TV world is going to have to throw away old notions and practices and really dig into the data.

Will that be you? Are you ready?

22 comments about "Why Have TV Viewers Stopped Channel Surfing?".
Check to receive email when comments are posted.
  1. Kevin Killion from Stone House Systems, Inc., November 16, 2017 at 5:12 p.m.

    This is good stuff.  I would love to know what metric you used against the AMRLD to quantify channel surfing.

  2. Paula Lynn from Who Else Unlimited, November 16, 2017 at 5:14 p.m.

    For one thing, the format that Comcast is using to find shows is not fun, it is disjointed. It wasn't as pretty last year and needed improvement, but it was easier. So we stay in the corners we know, once we find it which is a human characteristic. That part hasn't changed. Is 135 networks necessary and can enough people - outside of the top tier- earn or expect to earn enough income to stay in the industry ? So many questions.

  3. Ed Papazian from Media Dynamics Inc, November 16, 2017 at 5:34 p.m.

    Dave, this is very confusing. If you mean that the average viewer is watching fewer "linear TV" channels than before, can you quantify this with some Nielsen data? Also, if you are referring to broadcast TV as opposed to cable, I think that it is true that somewhat fewer channels are being sampled but what's the time frame? If it's weekly, of course but if it's monthly then I would expect any decline to be minimal. Finally, are you referring to primertime or all viewing, regardless of daypart? You raise some interesting questions but I think that many of us need a clearer explanation, coupled with tighter daypart and network type definitions and, if possible, some Nielsen backup data to come to grips with what you are saying.

  4. Dave Morgan from Simulmedia replied, November 16, 2017 at 6:37 p.m.

    Ed, Yes. I could have made it clearer. Also, they looked at the data in ways that are different than most TV viewing has been analyzed before, so the analysis isn't likely to line up the same way most TV viewing analysis has been done historically. The core data for this analysis was Nielsen's respondent data, the basis of the ratings. It includes both broadcast and cable networks. Fewer channels are consumed be each viewer in all time frames analyzed  - daily, weekly and monthly. The veiwing daypsrts analyzed were all day-parts, as well as time-shifted viewing. There is much less samping of viewing than we used to see. Folks seem to go more directly to what they want and stay there longer. We are creating a white paper on this with Forrester ... thanks so much for the suggestions. We'll work to address them in that paper.

  5. Kevin Killion from Stone House Systems, Inc., November 16, 2017 at 7:01 p.m.

    As a researcher of AMRLD since it became available, and MMRLD and both the aggregate and RLD versions of the PCS before that, I'm eager to hear more. This still needs quantifying, and identifying exactly each measurement used. For example, I'm not sure all persons reading this will agree on what they expect "channel surfing" to mean -- exactly -- or "less sampling of viewing".  I expect all that will be in your paper about this.

  6. Robert Barrows from R.M. Barrows, Inc. Advertising & Public Relations, November 16, 2017 at 7:19 p.m.

    Check out Springsteen's song "57 channels and nothing's on"

  7. Ed Papazian from Media Dynamics Inc, November 17, 2017 at 7:57 a.m.

    The contrasts between household usage and individual viewer channel choice patterns will be interesting. Nielsen usually shows household data---though I have seen some monthly data on viewers----and Nielsen has not, as far as I know, tried to break it down by daypart, demos, etc. ---unless this was done for or by a particular client and never publicized. The definitional question raised by Kevin is also important as well as the time frame. I hope to hear more about this, Dave, when your report is ready.

  8. Douglas Ferguson from College of Charleston, November 17, 2017 at 9:37 a.m.

    If "Netflix" is a channel, then it's bigger than any other channel offering high-quality shows, typically one at a time, serialized daily or weekly. Maybe binge viewing is the difference. Being able to find and watch anything on any device is a sea change for legacy media. Who needs to surf for something that probably isn't there when too much good stuff really is there 24/7.

  9. Ed Papazian from Media Dynamics Inc, November 17, 2017 at 10 a.m.

    Douglas, I assume that you noted the Nielsen report stating that the vast majority---80%---of SVOD viewing is to rerun fare not original content. I doubt that "binge viewing", which takes place sometimes---not always---for some people is the primary factor. The average Netflix subscriber devotes only 50 minutes per day to it's many episodes does that add up to on a "binge viewing" basis.

  10. Dave Morgan from Simulmedia replied, November 17, 2017 at 11:08 a.m.

    Ed, AMRLD now breaks out person level viewing data in addition to the household data, so its much easier to do the longitudanal analysis for dayparts, demos, etc. This is letting us all get a lot smarter in understanding how TV viewing is happening today, and how it's changing.

  11. Bob Smith from Kelliher Samets Volk, November 17, 2017 at 2:18 p.m.

    Wow!  This is extremely insightful. 

  12. Jack Wakshlag from Media Strategy, Research & Analytics, November 17, 2017 at 2:27 p.m.

    Always good and important to see some data, perhaps a small table, since going from 8 channels a week to 7 isn’t so much but if it’s now down to 4, that’s big. 

  13. Craig Mcdaniel from Sweepstakes Today LLC, November 17, 2017 at 3:35 p.m.

    Dave, nice job.  I think there is a different dynamic that has also taken place in the past twenty years.  This is consumer has only so many hours a day that they can use for entertainment. The total number of entertainment sources have increases exponentially. We have many more TV channels which is a given.  But before the internet there was VCR's then DVD's. Now comes the internet. We have gaming, social media, chat, texting and other sources for entertainment. Yet, the TV side has not really embraced the notation that the internet is a competitor of TV.  To take this a step further, many of my members enter sweepstakes during the commerical. They consider spending the time entering a sweep for a new car, trip or whatever more valueable than watching a TV commerical. 

    So if you want data, get it online instead?

  14. Ed Papazian from Media Dynamics Inc, November 17, 2017 at 5:33 p.m.

    Guys, before jumping to conclusions, let's see the data---if Dave will share it with us when the report is issued. While keeping an open mind, I would be surprised if a series of tabulations trending the number of channels that are viewed---including SVOD/OTT as well as digital venues like YouTube---as competition to "linear TV" ---showed a significant decline in the average adult's menu---say per week and/or month. If it's only "linear TV" and only over- the- air or cable channels that are referred to, I would expect to see fewer channels viewed per adult as alternate forms of TV viewing are competing for what was once a "linear TV " exclusive" ---or almost exclusive--- monopoly situation. I wonder how far back Nielsen has data on SVOD, ROKU, YouTube, etc. as the basis for such comparisons. But, as I said, let's see what the data tells us.

  15. Dave Morgan from Simulmedia replied, November 19, 2017 at 3:11 p.m.

    Jack, on a weekly basis, the number of channels the average viewer watches has dropped closer to 2 than 1. I'll make sure that the white paper lays it all out in detail. Stay tuned!

  16. Kevin Killion from Stone House Systems, Inc. replied, November 19, 2017 at 4:44 p.m.

    Paula has a very good point.  Searching for channels is a mess, so that has GOT to suppress number of channels watched.

    IMHO, one reason for declining shares of the big nets is that it has become so darn hard to find anything.

    In my home, we have Comcast cable with a TiVo box, so I'm not sure who to blame, but here goes.

    The big 4 (or so) networks are all listed by their local call letters, which puts them at the end ("W") of the long channel list.  You'd be crazy to choose to sort the channels by number since there is no discernible pattern.  (Why do viewers need channel numbers at all? We don't go to websites by typing in IP addresses.)

    Subchannels are intermixed with the main channels.  Network and subchannel names (ABC, CBS, Fox, Antenna, Me, etc.) are not shown so you better know the call and the subchannel number.

    SD and HD with the same content are intermixed with no easy way to suppress showing of a SD channel when an HD version is available; you have to remove them manually one by one.  There are no easy flags, icons or colors for HD; instead, the call letters are mangled, e.g., WBBMDT, but not in a consistent predictable way. Some HD-only channels have no designation at all of that status , while the only way to tell if an SD channel has an HD partner is to manually hunt the channel list, because sometimes that heretofore mentioned mangling upsets the natural alpha order.

    Some channels are just plain poorly identified: Headline News is in the list as, I'm not kidding, as "CNHNHD". Discovery is "TDCHD-E".

    Removing channels from the preferred list is a pure manual task and a royal pain. The screen enabling that is itself presented in SD, with only those mangled calls.  There is no way to quickly identify well-defined groups of channels you don't want, such as all foreign language channels, all sports channels, or all direct purchase channels. Periodically the lineups change, so that task has to be done again.  The other day I got around to do this again; it took an hour but I'm not quite done yet.

    So, what's on ABC, CBS, NBC tonight? Who knows!

  17. Ed Papazian from Media Dynamics Inc, November 19, 2017 at 6:25 p.m.

    Once again, I fear that we are talking mainly about primetime entertainment fare and really about the broadcast TV networks plus some cable channels, not the totality of TV viewing. The combined audience of ABC, CBS, Fox, NBC and the CW in primetime accounts for only  10-14% of the average adult's total TV viewing consumption. Add some of the cable channels which offer origina quality primetime fare and you probably raise the figure to 20-25%. As for the rest, a great deal of it remains habitual in nature. For example, if we watch early AM TV we know where our favorite shows is as well as several direct competitors. The same goes for the early news and the late night talk/variety shows and, of course the local station daytime, early and late news editions. I would also add many of the Monday-Friday daytime network and syndicated shows----"Judge Judy", "Dr Phil", a favorite soap, "Seinfeld" reruns, etc--- and this applies as well to the early evening celebrity newsmagazine  and game shows. It also is the case for sports, though here, I share the frustration of many trying to figure out which channel is carrying a favorite team today. Finally, taking basic cable, there are many thematic channels which are known to all of their fans---CNN, Fox News, MSNBC, The Weather Channel, The Food Channel, History Channel,and many others, so even though basic cable can be less predictable program scheduling-wise, fans of many genres have long since figured out where their favorite and near favorite shows are to be found.

    I wouldn't be at all surprised to see that the average adult now watches fewer broadcast channels in primetime than before, especially on a weekly basis, which would be a natural outcome of competition from SVOD/OTT, which are merely other channels for people to choose from, as well as whatever is attracted by digital venues and other venues. But I don't expect to see major changes affecting all of TV programming---especially the vast bulk of non-primetime content. Still, let's keep an open mind and see what the data says.

  18. John Collins from Media Storm, LLC, November 20, 2017 at 6:21 p.m.


    Good piece,  and from reading the comments you have the media math folks working hard on figuring out the numbers.   

    I am still waiting for the answer to the title of the piece.   "Why have TV viewers stopped channel surfing"  

    Could be that Channel surfing is an analog reference to how we used to find programming to watch,  we would scroll through the analog channels rapid fire, looking for something to catch our eye.  In today's digitally delivered environment,  you cannot channel surf like you used to.   

    In the digtial delivery system, each channel that you land on has to be delivered to your set top box,  that delivery takes milli-seconds longer than it used to,  and that delay has rendered good old channel surfing pointless.   

    Think back to your youth,  I am sure you had an early remote in your hand and would rapid fire change the TV channel and you could tell if you landed on anything worthwhile.   That rapid fire trial and error is mostly gone.   

    CTAM did an interesting study a few years back that should be revisited,  "How do you find content" Top 5 answers

    • Turn the TV On

    • Check your DVR

    • Check the onscreen TV Guide

    • Flip to a few of your favorite channels

    • Check the On Demand menu.

    I am sure that On Demand would now jump up the list.  

    Also do you remember the concept of LOP "least objectoinable program"  that is how we used to settle on content to watch.   I can live with this show, while I wait for the show I really want to watch comes on in 30 mintues.   those days are also gone.  

    In today's On Demand/DVR world,  you don't have to settle.  You can watch WHAT you want,  WHEN you want, on the device that is most convienent for your viewing circumstances.  that is what I call control.  

    To me that is WHY viewers have stopped channel surfing,  because they CAN.  

  19. Ed Papazian from Media Dynamics Inc, November 20, 2017 at 6:41 p.m.

    John, regarding "channel surfing", I assume that Dave is referring to looking around to try to find a worthwhile show to watch---mainly in primetime---again, my assumption---but not to dial switching, which is often a method of commercial avoidance. I pay close attention to this aspect in particular and I have not seen any signs of attrition in dial switching. TVision now tracks presence in the room and whether the "viewers" eyes are on the screen when program content is on the screen and when commercials appear, using a form of "eye camera". I wonder if they have any insights to give us as to the currnet incidence of dial switching during program content and/or commercials----like average minute activity, for example?

  20. Jeff Storan from Simulmedia, November 28, 2017 at 3:47 p.m.

    Here's a link to the companion white paper to Kyle Hubert's webinar, "The Hidden Story Behind TV's Ratings Decline."

    To address questions from column commenters on the measures and magnitude of the reduced channel flipping, in the white paper, we remark on the challenge of portraying channel flipping in AMRLD due to the “majority minute rule.”  If a respondent changes the channel multiple times within a single minute, AMRLD assigns the minute to the network that received the most of the viewer’s attention. AMRLD does not record sub-minute channel changes. 
    Sub-minute channel change events are knowable through analysis of second-by-second viewing data in Simulmedia’s set-top-box panel. While we don’t include the analysis in the white paper, we did discover in our STB data that channel changing within a viewing event is on the decline.
    To corroborate the STB data analysis showing reduced channel changing, we analyzed AMRLD to look at the number of networks the average TV viewer watches each day.
    In the webinar presentation, we noted that, in the period from 2013 to the present, the average number of networks watched live by each viewer each day has decreased from a little more than 6 to a little less than 5. The analysis included all broadcast and cable networks and covered viewing in all dayparts. 
    For the webinar presentation and the white paper, we also analyzed the share of viewers who watch different counts of networks live each day and found that the percentage of viewers who watch only 1, 2, or 3 networks has increased dramatically between 2013 and the present.

  21. Dave Morgan from Simulmedia replied, November 28, 2017 at 5:24 p.m.

    As you can see from the details in the white paper, I stand corrected on the size of the drop in the number of channels viewed (i misread the raw data ... much better to trust the work of the scientists than me ;-)
    It is down by one channel, not closer to 2 than 1 as i wrote in my comment before.
    White paper here:
    Please let me know what you think of it.

  22. Ed Papazian from Media Dynamics Inc, November 28, 2017 at 6:08 p.m.

    Dave, I've read the report and have several comments. First, I agree that viewing of live TV is down and, consequently, that the average number of  linear TV channels sampled would also be down. However, the base for this analysis is an average day, which is a very short time frame. Did you obtain any weekly or monthly data---which could alter the picture? Another question concerns the reason for the difference between the decline in total live viewing time and the relatively greater decline in average channel ratings. I would think that this is primarily a function of the increase in the number of channels available---which has risen dramatically over the past five years, especially, if you count alternative sources of content---Netflix, Hulu, etc. This change was not shown in the report---unless I missed it.

    It seems to me that if we confine ourselves to live, "linear TV", viewing, that there has been a relatively small but detectable change in short term channel usage and a modest increase in time spent per session----in the latter case, as I deduced from the charts, an increase of a bit more than two minutes per session ( about 15.3 minutes in 2014 to about 17.5 minutes in 2017--- as best as I can pinpoint it from the bar charts.

    One final point. You make the statement that thirty years ago an advertiser who bought a primetime spot on each of the three major networks on the same night would  reach 100% of all viewers while today the same buy would reach only 5%. Actually, a typical primetime "roadblock"---same ad on at the same time on all three networks--- probably could have been seen by about 30% of the TV population back in 1975-1980; the comparable figure today is more like 10%. So things aren't as bad as you indicated, where short term reach is concerned.

    I think that a longer term analysis would reveal that the typical adult samples about the same number of channels per week and month as before, however, the number of these that are "linear TV" channels is down while SVOD channels make up most of the difference. Also, I believe that were SVOD and options like Roku included in your analysis as well as delayed "linear TV" viewing, that this  would probably present a more balanced picture as the average viewer doesn't really care how we media folk characterize channels or how they are accessed. To the viewer, they are all options, not just some of them, and all of them vie for the viewer's attention, not just the "linear" ones.

    In any event, thank you for this analysis. It raises some interesting issues and may well serve as a base point for future studies of the changes in "linear TV" usage as it increasingly faces competition from alternative channels or methods of accessing TV content.

Next story loading loading..