Live TV Viewing Is Flat, DVR Viewing Is Up Slightly, And Online Viewing Is Still Tiny, Contrary To What You May Otherwise Have Heard

Categories: 1-Featured,TV Business,TV Ratings Reference

Written By

September 10th, 2013

retro tv television

If you spend too much time listening to some people you'd think nobody watches TV anymore, they all watch online, and TV viewing is dead.

For those folks, actual data is often inconvenient, like the numbers in this research note from Brian Wieser at Pivotal Research:

BOTTOM LINE: Yesterday, Nielsen published its "Cross-Platform" report for 2Q13. TV continues to hold up well - and dominate - among all audience groups, as it remains the broadest reaching medium with the greatest amount of inventory tonnage. The data illustrates why traditional TV will continue to be the primary beneficiary of ad budget allocations for large brands. This is favorable for companies under our coverage including CBS, Viacom and Discovery, but also highlights that Facebook and Google will need to focus on non-TV budgets to generate the growth that many investors expect from their online video initiatives.

Nielsen's Cross-Platform reports are key tools in assessing media consumption trends, as their processes reflect input from many of the industry's best researchers and decades of experience of stratifying their samples. The Cross-Platform report data is more meaningful than the ratings data much of the industry focuses on because this report is based upon more forms of TV consumption, including local and national viewing. By contrast, the NTR sample which most observers look at weekly or monthly is part of an incomplete sample for those looking to assess broader viewing trends. The Cross-Platform data does have limits, such as the exclusion of consumption on connected TVs and tablets. It is also limited by small sample sizes which mean that smaller audience groups may have large standard errors, leaving some room for interpretation of many of the figures produced in these reports. Historical comparisons may also be challenged by processing errors or other methodological changes.

Key points from our analysis of Nielsen's latest Cross-Platform report include the following:

·         Total time with "traditional" (i.e. live-only) TV was flat year-over-year, per person. We calculate nearly 11 billion person-hours of time-shifted viewing (i.e. via DVRs or VOD) during 2Q13. This amounted to 8.6% of TV viewing in the quarter up from 8.0% in the year-ago period. However, we calculate 114 billion person-hours of live TV viewing hours occurred during 2Q13, slightly higher than in 2Q12, meaning that in absolute numbers traditional TV held up despite a higher share of time-shifted viewing. We note these figures incorporate Nielsen's population estimates (which were flat year over year, vs. US Census Bureau estimates of 1% growth year-over-year).

·         Viewing of TV by kids 2-11 was flat per-person during 2Q13, although down in aggregate year-over-year largely because of Nielsen's reduced population estimates. During the 2011-2012 broadcast season, Nielsen incorporated an estimated population of 41.2 million kids 2-11, but during the 2012-13 season reduced that estimate by 4%. This has the effect of depressing reported kids' viewing levels to a notable degree, although the effect will rebound as Nielsen's population estimates will generally converge with those of the US Census Bureau.

·         Online video viewing reported by Nielsen equated to 2.3% of total TV viewing among the whole population, although unmeasured tablet viewing should add significantly to this amount. Among the population who actually watch online video content, the equivalent of 4.4% of viewing was reported to have occurred via online video. Among kids 2-11 only, the 1.7% of total viewing occurred via online video, with 3.1% of viewing among the sub-set that actually watched online video during the quarter. Historical data was characterized as not trend-able due to a processing error which rendered 2Q12 as inflated. However, we also note that these figures are likely understated given the absence of inclusion of tablets (although with only 27 million households - or around a quarter of the total - in possession of tablets as of 2Q13, our guess is that at best online video viewing might be double the stated figure).

·         Traditional TV remains relatively broadly popular, while online video and internet use in general remains concentrated among relatively smaller groups of the population. Nielsen's data indicates once again that while there is some concentration among TV viewers. The 16% of the population not on the internet paired with the 17% heaviest consumers of TV - one third of the population, in total - account for 64% of TV viewing. However, virtually the entire population continues to watch significant volumes of traditional TV (even the lightest 17% of TV viewers continue to watch 8x more traditional TV than they watch online video). Concentration among use in other media is much worse, and light users are very light users. For example, the 10% of the population which are the heaviest users of online video account for 84% of total online video viewing. Including the next 10% heaviest users, we can identify 20% of the population as accounting for 94% of online video consumption. Similarly, we can identify that less than 25% of the population account for nearly 90% of time spent with the internet more broadly.

As has long been the case, the data included in Nielsen's report highlights the nature of traditional TV as the medium best able to satisfy reach and frequency-based media goals against broad populations. This is important to note as web-based media owners such as Google, Facebook and others position themselves as capable of capturing traditional TV budgets. While there will always be certain marketers who look at the market through different lenses, or alternately get comfortable with the hybrid measurement processes and workflows that allow them to budget for and execute against video in a relatively holistic manner, our view is that these efforts will account for a small minority of video-based ad spending for the foreseeable future. Instead, we continue to believe that online video will primarily capture share from budgets which would otherwise be directed towards traditional premium display advertising (including conventional display inventory which would otherwise be sold by portals such as Yahoo, AOL and MSN).


  • Steve A

    So where’s the rest of the decline going if not to cable?

  • shogun

    Good points. There’s also the fact that Netflix’s content is mostly old shows and movies that have already made back there production costs on regular tv, so they can acquire it cheaper. If live TV dies out, Netflix would have to become a LOT more expensive, or we’d lose a ton of content.

  • Theoacme

    It’s amazing how little it can take to renew a show on cable, if it’s not one of the major cable channels – USA, TNT, etc.

    A show with an 0.59 P 25-54 (is that different than A 25/54?), Renovation Raiders, got renewed for a second season on HGTV, being the third highest rated prime time show in HGTV during its first season run, per HGTV press release.

    At 0.59 levels, even a few hundredths of a demo point can be materially useful to the income stream of HGTV, both in ad revenues, and retention on cable lineups (too low ratings, and they may be more vulnerable to being dropped).

    So, online viewing (including OnDemand type services), if they have the same commercial load during the C+3 period as on the cablecast, could eventually be worth it to some second-tier cable channels to pay for, and thus, eventually lead to its adoption by broadcast and top-tier cable channels.

  • Bill Gorman

    “So where’s the rest of the decline going if not to cable?”

    Everywhere else. Syndicated. Other dayparts besides primetime. Non-English broadcast.

  • Bill Gorman

    “P 25-54 (is that different than A 25/54?”

    No. P = people. A = adults.

    Usually see P associated with under 21 demos like 12-18.

  • Theoacme

    Non-English broadcast, yes…when Univision continues to occasionally lead nightly prime time broadcast demos (unthinkable ten years ago), and regularly beats more than the CW on some nights, TVBTN may have grown sufficiently by then to have to list their show-by-show demos along with the others…

    …and Telemundo is owned by NBC Universal Comcast, so if the parent chooses to add resources there, it could come to pass that Telemundo, too, may require show-by-show demo listings here. (Of course, by that time, the corpse of the CW may be under the knife of Dr. Morales in the morgue, cause of death the question – and the answer won’t be Frangelico iced cake :( )

  • Justin

    This report sounds highly inaccurate. For online viewing, did they include downloads? Streaming on other sites with third-party links such as Primewire? Or did they just count viewing on the actual network websites and Hulu?

    Game of Thrones, for example, averaged about 5 million viewers on TV but also averaged about 5 million illegal downloads. And that isn’t counting other streaming sources. While networks and studios don’t profit from these forms of viewing, accounting for them does represent a far more accurate measure of popularity.

  • Eugene

    I tried online viewing but found the quality to be abysmal. CBS’ video player was the worst of the bunch, but I had quality and performance problems with all of the network players, and ditto for Comedy Central. These networks should just upload the shows with commercials and allow you to stream in HD ala Netflix, and just remove the fast forward feature if they’re worried about commercials. I’ve had instances where the show never came back after the commercial, and your only option is to start the whole thing over….completely unacceptable with today’s technology, as Netflix proves every day. And so I’ve taken to frequenting free streaming websites that show no commercials, and shows are up approx 10 minutes after they go off air. Until the networks wake up and stream in HD quality and abandon their crappy players, give me one reason not to.

  • tjw


    Repeat after me:

    “Ratings are not a measure of popularity.”
    “Ratings are not a measure of popularity.”
    “Ratings are not a measure of popularity.”

  • Justin


    But the report isn’t about ratings. The report is analyzing total viewing figures across all platforms, which is a proxy for popularity. When people ask why traditional RATINGS are declining, the answer is that many people are turning to online VIEWING, which isn’t fully reflected in any measure of RATINGS. Perhaps online viewing is only 3% of RATINGS, but it is a much, much higher percentage of total VIEWING.

  • tjw

    Well, this includes television watched online, though it probably doesn’t count illegal viewing (like torrenting). But that’s such a tiny piece of the puzzle. Game of Thrones was illegally downloaded 5 million times TOTAL for this past season, and at least half of that was from overseas. Illegal downloads might count for a few million viewing hours every month but Americans watch on the order of 10 billion hours of television every month, so you’re talking about at most maybe .1% of total viewing.

    Just because data doesn’t conform to your worldview doesn’t mean it’s “highly inaccurate.”

  • Justin

    Wrong. Just the season premiere episode alone of Game of Thrones was downloaded by 5.2 million people on just Torrent, and not even including other sites for downloading. You are right though that overseas accounts for a large percentage of this.

    The data doesn’t conform to my worldview because the way in which it is measured is imprecise and inaccurate. Counting every form of online viewing across all platforms would yield significantly more accurate results. Simply measuring the number of views on the networks’ online video players is a paltry effort.

  • tjw

    You’re right about Game of Thrones. I read my source wrong. But you’re still looking at tiny numbers when you’re talking in the aggregate. Even if every hour of every night of broadcast primetime was downloaded by a half million Americans (which they’re not), that would still account for less than .5% of total viewing.

    The data measures precisely what it’s intended to measure: how people legally view media. Whether that’s on television, online, on their phones, however. What value is there in counting people who don’t pay for their content (either with money or ad-playing) other than to allow Game of Thrones fans to pat each other on the back?

    If you want to really dig into the problem areas, it’d be nice if Nielsen broke out DVD/Blu Ray player and Game Console data so we’d know exactly how often each is being for Netflix/Hulu/Amazon viewing versus their traditional purpose, though I still think there wouldn’t be as much as most people think.

  • tested

    I took a quick glance at all this and had two quick thoughts:
    1. This report was not done to inform the general public about current trends in television viewing. It was done for businesses to have a better understanding of what they’re getting for their advertising dollar and where they should spend their money.
    2. Nielsen’s report likely excludes a lot of Netflix viewing – and that should be a big relief to cable and broadcast networks. This report shows regular TV viewing is not taking a big drop because of Netflix.

    I don’t think Nielsen is a very good system.. but it’s the only company doing this kind of data collection. (RenTrak is just too new and isn’t doing a lot of reports like this that are publicly available) I think Nielsen needs to adjust their definition of television viewing, vastly expand their sample of viewers and improve their ability to measure viewing on mobile devices.

  • James

    So strange.

    I know so many people who are what they watch is 90% netflix or equivalent – It’s very hard to believe the nielsen rating on that demographic .

© 2015 Tribune Digital Ventures