How much has TV viewing really changed?

I was listening to the Rule of Three podcast with Laurence Rickard (of Horrible Histories and Yonderland fame, to anyone with young children in the UK), and there is some conversation about producing TV programmes for different channels (CBBC- Horrible Histories, Sky- Yonderland, and BBC1- Ghosts), and the differences in audience expectations for a programme that goes out on a given channel, at a particular time - and how changing viewing behaviours mean that how programmes are watched is not as closely tied to how they are scheduled any more.

Joel Morris (one of the podcast hosts) says, about 4 and a half minutes in;

"We did some work about 5 years ago for Gogglebox [...] I was always under the impression that Googlebox was a fiction, that families didn't gather around the TV and watch it anymore [...], and we looked up the BARB figures (about 5 years ago) were that 85% of programmes were watched live, by a family, on a sofa, when they went out. And I thought - god, I thought that was a vintage and antique thing. I am sure that is not true any more, and that must have changed really quickly, and I'm wondering whether the culture of comedy, where the demand is for things to hit straight away, whether people have quite caught up with the fact that people consume everything so differently now."

I'm really interested in changing TV viewing behaviour for all sorts of reasons (some of which I talked about here- an old post, but still relevant), so I thought I'd take a look at whether this is true - what has changed in the last 5 years?

First of all, for the sake of clarity, lets look at what is being said;

85% of programmes were watched live, by a family, on a sofa, when they went out.

OK- there's a few things going on there, so for the sake of clarification;

• "85% of programmes" - this is a bit of a fuzzy one, because you aren't really talking about "85% of programmes" - a given programme can be watched live, by a family, on a sofa, when it goes out by one household, but watched from a recording, by a single person, in a bedroom, the day after it goes out next door. I'm going to assume that what Morris meant was "85% of viewing time to programmes" - ie. "85% of all the time being spent by people watching television programmes". Because nobody really cares about "percent of all programmes" - a metric which would put a world cup final on equal footing to an obscure programme on an obscure digital channel, broadcast at 3am on a Tuesday...
• "watched live" and "when they went out" - I'm going to lump those together as meaning "watched when they are broadcast", because "live" has a slightly different meaning. (ie. Gogglebox is pre-recorded, so its never "watched live"... unless you are using the TV Licence definition of "live TV"...) So I'm looking at what is viewed when it was broadcast, vs what is timeshifted (ie. watched at a different time to when it was broadcast, whether recorded, streamed or downloaded.)
• "by a family" - BARB data doesn't really have an easy way to say what is watched "by a family", but I can look at what is watched by an individual on their own vs. "shared viewing" watched by 2 or more people together.
• "on a sofa" - OK, this isn't really something that is captured by the BARB data - you can look at what room a TV is in (ie. living room, bedroom, kitchen etc.), but I'm interpreting this to mean "on a traditional TV set" - as opposed to on a computer, tablet or smartphone.

Luckily for me, the BARB website has some of the data available to the public.

But there's another bit of the quote worth pulling out;

Firstly, live vs. catch-up viewing; in 2014, live viewing accounted for between 85.8% (w/c December 22nd) and 90% (w/c June 23rd). In 2019, that figure has changed... a little. The most recent data at the time of writing is for w/c 1st July, where live viewing accounted for 85.8% of viewing- the highest for the year to date. (The lowest was 83.2%, w/c 8th April.) I suspect that this live vs. recorded viewing is what he was talking about - just because the number is so close. But has it changed much? I'd say no. A bit, but not much.

Next is "by a family" - which I'm looking at as viewing by one person vs viewing by more than one person. This is a bit trickier to get hold of the numbers - I couldn't find anything on the BARB website for solo vs shared viewing, but as I work for a BARB subscriber, I do have access to their full data. This is what the share of "solus" viewing by month looks like for the last 7 years;

Solus as a share of total viewing minutes

So - 85% of viewing is definitely not watched "as a family", because nearly half is watched alone... but that said, the point here is that I'm not seeing anything here that makes me think that there has been a significant change in viewing behaviour over the last decade. (Interestingly, we watch more TV on our own in the summer months- which is when we tend to watch less TV overall.)

Finally, "on a sofa"... or, my interpretation, on different screens.

BARB report two sets of data at the moment - firstly, whats watched on TV screens, based on a panel of households with meters monitoring what the TV is doing. Secondly, what is being watched by the "players" (ie. iPlayer, ITV Hub, All 4, Sky Go etc.) on other devices, based on census-level data from the players themselves.¹

You can see the time by device data on the BARB website here - I've had to pull the equivalent data from BARB myself to get to this chart of total viewing minutes.

From the bottom, we have a combination of "live" (ie. as broadcast) viewing and VOSDAL - Viewing Same Day As Live (ie. watched on the same day as the broadcast). Then we have viewing that is watched between 1-7 days after it was broadcast, and a much smaller slice of programming watched between 8-28 days after it was broadcast.

Then there's a purple slice of viewing that is "unknown" - the way BARB works is by matching what is on a TV with an archive of broadcast content from the last 28 days - anything that isn't matched is "unknown"; so that will cover watching DVDs, or watching Netflix exclusives, or playing computer games. Its generally understood that most of this is accounted for by Netflix and Amazon Prime, but exactly how much isn't massively clear - and certainly a bigger topic than I can cover here (where I'm already oversimplifying more than I would like...)

Then there is a tiny sliver along the top that is the billion or so weekly minutes of "player" TV being sent out to computers, tablets and smartphones. (ie. very unlikely to be watched by a family, together, on a sofa.) A billion sounds like a big number out of context...

As I said, I've been watching these numbers for quite a while, so none of this really surprised me very much. About 12 years ago, a colleague of mine did some analysis on PVR viewing where he found that when a household gets a PVR, they do start skipping adverts - but they also watch more TV, and the extra TV viewing outweighs the ad skipping, so they actually end up seeing more advertising - at least, for the first 6 months or so. That taught me two important lessons;

• First, we are incredibly bad at understanding our own behaviour. (I wrote about this some years ago - see "Do you really know how much TV you watch?")
• Second, the intuitive conclusion - in this case, we see less adverts when we start skipping them - isn't necessarily backed up by the data.

Why the Sofa is the most important media device

My theory is that the space that "TV" fits into people's lives simply hasn't really changed much. Generally speaking, people go to work, come home, deal with the household chores that have to be dealt with, and then at the end of the day they sit down on the most comfortable chair in the house - which has a TV in front of it - and they watch TV.

The thing in the living room with the biggest influence on TV viewing isn't the TV (big, small, smart, dumb etc.) or the things plugged into it (ie. the satellite/cable box, recorder, DVD player, games consoles, wifi dongle etc.), or the things that are on the screen (TV broadcast, videos, DVDs, Netflix, games etc.), but the sofa.

As Marshall McLuhan said, "The medium is the massage". Or, "Content is the juicy piece of meat that distracts the watcdog of the mind". If you want to see the effect that TV has on people, watch the people - not the TVs.

That said, exactly what "watch TV" means is changing - I grew up with 4 channels to choose from, which meant that the best way of representing how people watch TV was the idea of "least objectionable programming" - people would see what the options were and settle on the one they objected to the least; the least boring, least offensive, least upsetting or unsuitable option would get watched. That model doesn't really work when you've not just got a dozen or so channels to choose from (just looking at the channel repertoire of what people actually watch - ignoring the dozens and dozens of channels in the EPG that you never actually switch over to - generally, people have about 23 channels that they actually watch, no matter how many channels they can watch), but also a library of pre-recorded favourites on your Sky+/PVR (sure, we had VHS tapes along with our 4 channels, but actually choosing one and watching it was a very different experience), as well as the archive of programmes available on the iPlayer/ITV Hub/All 4/Netflix/Amazon Prime/YouTube etc.

But our perception changes much faster than our behaviour, for a number of reasons;

• The more we think about something, the more we remember it. That means if we are looking through a library of content - whether actively looking for a specific programme or browsing through genres hoping to discover something appealing - we are much more likely to remember that act than flicking through channels to see what is on - which in turn, we are much more likely to remember than watching whatever happened to be after the programme we were watching because nobody objected enough to switch over. (Which is also related to why we remember fast-forwarding through the adverts more than we remember not fast-forwarding through the adverts, and similarly over-estimate how much we do it.) Our perception of ourselves is fundamentally biased. (Similarly, we remember doing stuff on our phone while watching TV more than we remember not doing stuff while watching TV.)
• We focus on the new - what has changed - and ignore the familiar until it goes away. So when we think about how TV is changing, we think about the technology drivers - big screens, Sky+, iPlayer, iPads, Netflix - and don't think about what is stopping things from changing. I suspect that if you were to take away the sofa from a typical household - or even to move it into a different room - you would see bigger changes in behaviour than introducing any piece of technology. (Well, you'd probably just see people moving it back again, but you get the idea...)
• We lie to ourselves. We like to think that we are the sort of person who doesn't just spend hours watching TV every single day. We like to think we are the sort of person who carefully curates and selects the content that is worth our time. (This is on top of the fact that we are wrong about how we are spending our time in the first place.)
• We are lied to. What the media tells us about how everyone else is watching TV is about the changes - advertising the new technology and the new services, reporting on the impact of the new technology (and the PR that promotes it), the science-fiction vision of how we will watch TV in the future - all of which builds a false picture in our minds of how 'everyone' watches TV, which we tend to assume applies to ourselves as well. I would put the podcast that prompted this post into this category; we hear people talking about how everything is so different now, so - unless we happen to be paying particular attention to the particular area - why wouldn't we believe it? (At the end of the day, that is a big part of why TV advertising works so well.)
• We lie to others. We don't like to present ourselves as the sort of person who spends hours indiscriminately watching TV every single day. We like to present ourselves as someone who is in touch with the current trends, using the latest technology. (This is on top of the lie we tell ourselves, on top of being wrong about it in the first place.) If a doctor asks a 25 year old how much alcohol they drink every week, they probably don't expect the answer to be accurate. (But they probably do expect the person they are asking to think about how much they actually drink, versus how much they should drink, and how big a lie they feel they should respond with...)

So, in conclusion, I don't think "TV viewing" has really changed all that much - in the sense of the way that we watch TV. The television industry is changing enormously - what we choose to watch, how we choose it, how we watch it, what is available to us, who is paying for it, how it gets comissioned etc. etc. All of which is a really interesting space to watch.

But the thing that remains really interesting to me - the basic human behaviour of sitting down on the most comfortable chair in the house at the end of the day and watching "audio-visual content" (for want of a better word) on the best screen in the house is a behaviour that has remained remarkably consistent for a remarkably long time.

1. The methodology is a bit too dull to go into in detail, but in a nutshell - each video has a tag which fires when it is played, telling BARB that it is playing. So BARB get a record of all viewing by the players - not just based on the panel, but based on every single view. What they don't have is a way to say who is watching - ie. old, young, male, female etc. - or how many people are watching. So technically, it isn't comparable data - but its the closest that is available. If you don't work in media research, you probably don't care. If you do, you probably know all this already. ↩