Ok, the Nielsen Cross Platform quarterly report is out, and we find it hard not to whine immediately about its inaccuracy. It is, in fact, perfectly accurate, it’s just not easy to benefit from.
The theme for this report is a look across all screens, but it’s as if someone was hamstrung into making TV viewing the best statistic possible. If you believe what Nielsen appears to be saying, you’d agree that pay TV operators were the richest in the world, because people are hardly ever on the internet, rarely watch DVD or DVR content and absolutely never stream video on any device other than a TV.
But Nielsen says very clearly that these days all of its calculations are done as a percentage of all Americans over the age of two, so you have to unravel the calculation by dividing by them by the 298 million Americans that includes, and then multiplying back out by the number of people who have a device, in order to make any sense.
Take its first graph about the average time spent on a variety of activities each day. In Q1 2013 Nielsen says that the average American watches 4 hours 39 minutes of TV, one minute more than last year and about ten minutes less than 2010. They watch 26 minutes only of DVR playback, play 13 minutes of games, and watch DVDs for 11 minutes. Now let’s re-interpret that.
Everyone knows that is rubbish, because we all watch ALL of our TV without viewing adverts, on a DVR, either watching it the same night, or a few days later, speeding through the adverts. We know this, but we are perfectly happy to believe that we are the only people doing this, so we keep quiet about it.
These figures stack up because any DVR viewing that is watched within 7 days of a TV program counts as watching it live in the numbers that Nielsen’s gives to its advertisers and in a way this is right. If you watched it WITH the adverts, then it’s probably in the same period that your campaign was scheduled for, but of course the reason we watch it after the live broadcast is to fast forward through the adverts.
There is a lot of speculation about how the DVR is really used. Some make the point that more TV is watched Live during Prime time, because the programming is so good that people want to watch it as soon as possible and use the adverts to catch up on what their friends thought of it, on the social networks. Well that’s optimistic, but if we look outside of Nielsen, there are some other estimates of how much time shifted viewing which goes on during that 7 day period. Most estimates vary anywhere from 14% of TV up to 24% of it.
So the first thing we might do is pick a figure in between, say 20% and take those 4 hours 39 minutes a day and add 20% of it to the DVR pile, around 56 minutes. Suddenly we have something like 1 hour 22 minutes of DVR time, but that’s not quite right either.
This is because there are around 51.7 million home in the US, out of some 112 million that have DVRs, so that makes only 46% of these homes are viewing that extra 56 minutes.
And actually the 26 minutes that Nielsen reports has already been divided by 298 million people, rather than the number of people that have a DVR, around 46% of that 298 million. So multiply it back out and you have 26 times 298 million minute, and divide it back again and you have 137 million people that had access to a DVR and so they must have consumed all of those multiple 26 minute chunks of DVR time. If you reverse that calculation you find that once again this is about 56 minutes of TV each day when people with a DVR engaged in time shifted viewing outside the 7 days, or roughly about 7.7 billion minutes across 137 million people.
It’s still not right. That’s because US homes have multiple TVs in them, so we are assuming that everyone in those homes has access to a TV that has a DVR on it, but that’s a bridge too far, let’s assume this is more or less right since multiple people can be time shifting on each device.
If you add to that the 56 minutes of time shifted viewing each day that Nielsen is not counting, that comes to 1 hour 52 minutes of programming each day being watched time shifted per day for all people in a DVR home.
The reason that Nielsen says it doesn’t actually record this is because DVR, and start over and catch up and VoD are all being lumped in, so it says it cannot calculate this. It can, it just doesn’t want to.
But it’s still not right. Those 56 minute lumps are the broadcast lengths of the programs that people watch. In actual fact about 22 minutes per hour were adverts, and assuming that these are skipped fairly expertly, that means for each hour that Nielsen says we watch time shifted TV, we only actually watch 1 hour less 22 minutes of advertising , so 38 minutes.
Perhaps it would be best to say that people in DVR homes watch roughly two one hours programs on the DVR each day, but speed through the adverts and manage to do this in about 76 minutes. So in a DVR home they watch 2 hours and 46 minutes of live TV, and about 76 minutes of time shifted TV with no advertising. While in a house with no DVR they watch about 5 hours 25 of TV, of which precisely 2 hours is advertising. This is the ticking time bomb the Nielsen is trying to keep its customers from understanding.
And what about OTT viewing time. It says that the average smartphone user only streams video for about 1 hour 15 minutes each month. That comes to 2.5 minutes a day. Tablet owners Nielsen says stream video for 1 hour 48 minutes a month, or around 3.6 minutes a day. Of course this is nonsense, and the same principle applies here. We know that our children go upstairs and watch about 2 hours of TV on their tablet after bed time, so where is that recorded?
It’s tough to say, Nielsen still offers streaming calculations based only those who stream, and it does the same for mobile, where any web, subscription or App video usage are included. As far as we can work out it does not then divide the minutes by the entire population of the US, but to get such low figures you would think it has to.
Nielsen even breaks out the top quartile of streaming users and says that each day these advanced users, on average spend 26.8 minutes streaming video from the internet, a further 62.3 minutes just using the internet, but 243 minutes on the TV, each day. We assume that the video and internet usage that these people do at work doesn’t count, but even so, less than 30 minutes streaming compared to four hours TV, seems very unlike any usage profile that we can intuitively imagine.
There just isn’t enough information available to tell how Nielsen has twisted this particular set of figures, but we assume that it has something to do with that perverse process Nielsen puts the DVR through. Given that around 50% of people in the US have smart phones and that less than half of those subscribe to a video service like their cable TV Everywhere service, than it could just be spreading the long streaming hours of some viewers over those who do not have a phone, and much the same could be happening with tablets, starting from an even lower population.
As far as OTT is concerned, or streaming or DVR usage or anything new in TV, Nielsen is an enemy of the state, with a huge vested interest to get it wrong. The disturbing thing is that as we tried to find other examples of other interpretations of the data, we come across an enormous list of publications and blogs simply re-iterating the Nielsen party line. These writers must be wondering if perhaps they are the only people using video streaming or fast forwarding through adverts. No you are not, everyone is doing it, it’s just that nobody’s counting.
But Nielsen does have this data, and it must make worrying reading seeing how habits are beginning to change, and how that change puts everything that Nielsen stands for in jeopardy.
Just to finish we thought we’d quote the author of the report, to show you how far they are removed from reality, ‘In looking at recent viewing trends across screens, one thing remains constant: the amount of traditional TV and digital content we consume is increasing. It just goes to show that no matter how busy our lives might get, we always seem to find the time to watch the content that appeals to us.’
‘..traditional TV remains vibrant and continues to thrive as viewers continue to engage with their sets by seeking out the entertainment and information that appeals to them. In fact, traditional TV viewing has grown year-over-year among the total US population. Hallelujah, and Amen.