You are on page 1of 21

Our annual exploration of the UK’s viewing habits

MAY 2018
bjective information is critical
for the UK’s television and
advertising industry. Each year,
£7.5 billion is spent by broadcasters
and advertisers on the production
and distribution of programme and
commercial content. BARB provides
an independent, joint-industry
currency that is trusted to assess
the return on this investment.

Dealing with audience fragmentation

has been a constant aspect of
BARB’s work since our formation
in 1981. Fragmentation today is
driven by the rise of programmes
and commercials being delivered
online and on-demand.

We constantly develop our

measurement techniques to meet
the needs of the industry with
high-quality data. In recent years
we have developed techniques
for tracking live and on-demand
audiences across TV sets, tablets,
PCs and smartphones. We have
also pioneered the reporting of
dynamically inserted advertising.

The Viewing Report brings to life

the latest insights from BARB.
We hope you enjoy reading it.

Tim Harford
Tim is an author, columnist
for the Financial Times


A guide to statistics in a misleading age
Tim Harford
and presenter of Radio 4’s More 9 A matter of fact
I PA & I S B A
or Less. In an article originally Welcome to the 2018 edition of The Viewing people watching programmes across four BARB NOW HAS
published in the Financial Times, Report. screens. In BARB Explained, you can read
Tim offers a guide to statistics in a
10 W h a t w i l l v i d e o m e a s u r e m e n t l o o k
how we are delivering a trusted source of
misleading age.
like in 2022?
When I wrote my introduction for last year's viewing behaviour across TV sets, tablets, PCs TO REPORT
Ed Shedd
edition, the speech given by Marc Pritchard, and smartphones. AUDIENCES
Ed Shedd 12 G e t t i n g ‘ e x t r a ’ o v e r B A R B Procter & Gamble's global marketing chief, FOR ONLINE
Ed Shedd leads Deloitte’s Neil Mortensen was still ringing in the ears of the industry. It In a guest essay, ITV’s Neil Mortensen uses
North West Europe was the catalyst for many things, not least a Love Island as a case study to demonstrate
T V, O N - D E M A N D
Technology, Media & Telecoms (TMT) NEW INSIGHT renewed scrutiny on audience measurement. why a trusted currency is still relevant in PROGRAMMES AND
industry team, as well as the UK We like scrutiny at BARB; it's part of our DNA. an era of proliferating data sources. BARB D Y N A M I C A L LY-
TMT industry team. He looks into viewing figures are the catalyst for blending SERVED
the future to ask: what kind of video 14 T r a c k i n g n o n - l i n e a r v i e w i n g Here in the UK, the IPA and ISBA published insight from different sources in a way that

measurement will the industry need A Matter of Fact which reinforced the amplifies the strengths of each dataset. ADVERTISING
16 B r i n g o n t h e b o x s e t s
in 2022? importance of having accountable media
20 To p o n - d e m a n d p r o g r a m m e s audience data. It emphasised how the Deloitte’s Ed Shedd looks further ahead in his
Neil Mortensen Online TV viewing principles that underpin BARB and other essay and considers what our industry will
Neil Mortensen is Director joint industry currencies are the best-in- need in 2022. Not all data are equal and we
of Audiences, ITV, where he CENTREFOLD - class approach; independence, objectivity face choices on what to measure, from screen
has responsibility for ITV’s research and transparency are enshrined in these size and audio quality to video visibility,
and planning teams and sits on the
THE TOP TENS OF 2017* principles. prominence and adjacent content.
BARB board. Neil gets ‘extra’ on
BARB, discussing the importance 21 To p l i v e p r o g r a m m e s BARB has delivered a trusted currency since Returning to the need for trusted data, we

of the data for broadcasters and Online TV viewing 1981 and is committed to delivering relevant feature Tim Harford’s guide to statistics in
explaining why the future is bright audience measurement in an ever-changing a misleading age. You might know Tim as
22 O n l i n e T V v i e w i n g p a r t o n e :
for audience insight. world. This is why we now have techniques the Undercover Economist; this article first
On-demand v live streaming
to report audiences for online TV, on-demand appeared in the Financial Times. He provides
24 O n l i n e T V v i e w i n g p a r t t w o : programmes and dynamically-served practical tips on how to decipher the barrage
PC, tablet or smartphone? advertising. of statistical propaganda that increasingly
fills our lives. Tim wasn't writing with
26 O n l i n e T V v i e w i n g p a r t t h r e e : You'll find a bucketful of new insight from our audience measurement in mind, although his
W h o i s d o i n g t h e v i e w i n g? new techniques in this annual report. advice is very pertinent in our world.

28 S V O D n a t i o n We bring to life how on-demand viewing There's much for you to read, so let me get

drove audience levels for programmes out of your way.
BARB EXPLAINED such as The Tunnel: Vengeance and Top of
the Lake, both before and after they were Happy reading.
30 D e l v i n g i n t o d i f f e r e n t a s p e c t s o f broadcast as part of a linear schedule. We
what BARB does showcase the most watched non-linear
programmes on TV sets, and also the
U P DAT E D I N S I G H T programmes that were most popular on
tablets, PCs and smartphones. We look
at when people watch online TV on these
34  I n t r o d u c t i o n
devices and find out who is watching,
35  T i m e s h i f t taking The Great British Bake Off final as an
example. And we keep an eye on the growth
36 S h a r e b y g e n r e of SVOD services such as Netflix, Amazon
Prime Video and Now TV.
37 S h a r e b y b r o a d c a s t e r
2018 is the year we will deliver the next stage
38 T r a i n i n g & f i n d o u t m o r e of Project Dovetail. Building on the successful Justin Sampson
beta phase that reports the number of Chief Executive
* Please let us know if the centrefold has already devices being used to watch online TV BARB
b e e n r e m o v e d a n d y o u w o u l d l i k e a n o t h e r. programmes, we will publish the number of

“The best financial advice for most “chocolate prevents dementia”. If we don’t
people would fit on an index card.” That’s notice and pay attention to those feelings,
the gist of an off-hand comment in 2013 we’re off to a shaky start.
by Harold Pollack, a professor at the
University of Chicago. What sort of feelings? Defensiveness.
Triumphalism. Righteous anger. Evangelical
When I heard about Pollack’s notion, I asked fervour. Or, when it comes to chocolate and
myself: would this work for statistics, too? dementia, relief. It’s fine to have an emotional
There are some obvious parallels. In each response to a chart or shocking statistic - but
case, common sense goes a surprisingly long we should not be led astray by it.
way; in each case, dizzying numbers and
impenetrable jargon loom; in each case, there There are certain claims we rush to tell the
are stubborn technical details that matter; world, others we use to rally like-minded
and, in each case, there are people with a people, still others we refuse to believe.
sharp incentive to lead us astray. Our belief or disbelief in these claims is part
of who we feel we are. If we don’t at least
The case for everyday practical numeracy acknowledge that we may be bringing some
has never been more urgent. Statistical emotional baggage along with us, we have
claims fill our newspapers and social media little chance of discerning what’s true.
feeds, unfiltered by expert judgment and
often designed as a political weapon. We do The second crucial piece of advice is to
not necessarily trust the experts - or more understand the claim. That seems obvious.
precisely, we may have our own distinctive But all too often we leap to disbelieve or figuring out which one is being made before We should never forget, either, that statistics
view of who counts as an expert and who believe (and repeat) a claim without pausing retweeting it. are a summary of a more complicated truth.
does not. to ask whether we really understand what For example, what’s happening to wages? We
the claim is. To quote Douglas Adams’s All this can seem like pedantry - or worse, might look at the median increase in wages,
Nor are we passive consumers of statistical philosophical supercomputer, Deep Thought, a cynical attempt to muddy the waters which isn’t the same thing as the increase
propaganda; we are the medium through “once you know what the question actually and suggest that you can prove anything in the median wage - not at all. Sir Andrew
which the propaganda spreads. We are is, you’ll know what the answer means”. with statistics. But there is little point Dilnot, former chair of the UK Statistics
arbiters of what others will see: what we in trying to evaluate whether a claim is Authority, warns that an average can never
retweet, like or share online determines For example, take the claim “inequality true if one is unclear what the claim convey the whole of a complex story. “It’s
whether a claim goes viral or vanishes. If we of income before taxes is rising” (and you even means. like trying to see what’s in a room by peering
fall for lies, we become unwittingly complicit should be asking yourself, since when?), there through the keyhole,” he tells me.
in deceiving others. On the bright side, we are several different ways to measure this. One particular obstacle to our
have more tools than ever to help weigh up One approach is to compare the income of understanding is the question of In short, “you need to ask yourself what’s
what we see before we share it - if we are able people at the 90th percentile and the 10th causation. “Taller children have a higher being left out,” says Mona Chalabi, data
and willing to use them. percentile, but that tells us nothing about reading age”, goes the headline. This may editor for The Guardian US. There is no
the super-rich, nor the ordinary people in summarise the results of a careful study shame in leaving something out. No chart,
My statistical postcard begins with advice the middle. An alternative is to examine about nutrition and cognition. Or it may table or tweet can contain everything. But
about emotion rather than logic. When you the income share of the top 1 per cent - but simply reflect the obvious point that eight- what is missing can matter.
I F W E FA L L FO R encounter a new statistical claim, observe this approach has the opposite weakness, year-olds read better than four-year-olds
LIES, WE BECOME your feelings. Yes, it sounds like a line from telling us nothing about how the poorest fare - and are taller. Causation is philosophically Channel the spirit of film noir: get the
U N W I T T I N G LY Star Wars, but we rarely believe anything relative to the majority. and technically a knotty business but, for backstory. Of all the statistical claims in the
because we’re compelled to do so by pure the casual consumer of statistics, the world, this particular stat fatale appeared in
COMPLICIT deduction or irrefutable evidence. We have There is no single right answer. In fact, question is not so complicated: just ask your newspaper or social media feed, dressed
IN DECEIVING feelings about many of the claims we might there are many true statements that one whether a causal claim is being made, and to impress. Why? Where did it come from?
OTHERS read - anything from “inequality is rising” to can make about inequality. It may be worth whether it might be justified. Why are you seeing it?

Sometimes the answer is little short of a “It is better to be vaguely right than exactly CURIOSITY IS
conspiracy: a PR company wanted to sell ice wrong”, wrote Carveth Read in Logic (1898),
cream, so paid a penny-ante academic to put and excessive precision can lead people
together the equation for the perfect summer astray. It makes numbers needlessly VIRTUE BECAUSE

afternoon, pushed out a press release on a cumbersome to remember and to handle. So, IT ENCOURAGES
quiet news day, and won attention in a media embrace imprecision. The budget of the NHS US TO WORK A
environment hungry for clicks. in the UK is about £10bn a month. One can be
much more precise about these things, but LITTLE HARDER
Just as often, the answer is innocent but carrying the approximate number around in TO U N D E RSTA N D
unedifying: publication bias. A study my head lets me judge pretty quickly when - W H AT W E A R E
confirming what we already knew is unlikely say - a £50m spending boost is noteworthy,
to make news. But a study with a surprising or a rounding error.
result - maybe smoking doesn’t cause cancer
after all - is worth a headline. The new study Be curious. Curiosity is a cardinal virtue The IPA and ISBA joined together in June on trust. Proven media metrics are vital in
may have been rigorously conducted but is because it encourages us to work a little 2017 to call upon the media industry demonstrating the effectiveness of any
probably wrong: one must weigh it up against harder to understand what we are being told, to demand and provide objective and media channel.
decades of contrary evidence. and to enjoy the surprises along the way. independent data to underpin the
accountability of all media and their The principle of joint industry oversight
Publication bias is an even bigger problem in This is partly because almost any statistical respective trading markets. of the research underpinning media is
the media – and perhaps bigger yet in social statement raises questions: who claims this? fundamental to confidence in advertising.
media. Increasingly, we see a statistical claim Why? What does this number mean? What’s A Matter of Fact highlighted that the long- Joint Industry Currencies (JICs) have provided
because people like us thought it was worth a missing? If a statistic is worth sharing, isn’t it term health and prosperity of the media the advertising industry with objective inputs
Like on Facebook. worth understanding first? The digital age is industry is reliant on objectivity and into the calculation of marketing effectiveness
full of informational snares - but it also makes confidence. This can only be maintained for many years; their metrics are derived from
OK. You’ve noted your own emotions, it easier to look a little deeper before our if trading data is understood and trusted. independent data-collection methodologies.
checked the backstory and understood minds snap shut on an answer. Audience data is a cornerstone of the
the claim being made. Now you need to industry; it is fundamental to decision- The JICs are increasingly moving to hybrid
put things in perspective. A few months While curiosity gives us the motivation to ask making and monetary transactions. measurement systems that combine proven,
ago, a horrified citizen asked me on Twitter another question, it gives us something else, people-based research techniques with
whether it could be true that in the UK, too: a willingness to change our minds. If we Media data needs to deliver a trusted connected online data. It is vital that the
seven million disposable coffee cups were treat statistical surprises as mysteries to be assessment of the size and composition of a industry is reassured about the quality
thrown away every day. I didn’t have resolved, we are more likely to spot statistical campaign’s audience. This is vital to calculate and veracity of any proprietary data that
an answer. But I did have an alternative foul play, but we are also more open-minded a campaign’s return on investment, whether is absorbed into these industry-owned
question: is that a big number? The when faced with rigorous new evidence. the objective is driving sales, building brand currencies.
population of the UK is 65 million. If one awareness or any other goals.
person in 10 used a disposable cup each day, Isaac Asimov is thought to have said, “The The IPA and ISBA have called on their
that would do the job. most exciting phrase in science isn’t ‘Eureka!’, The need for consistent measurement of respective members to uphold the following
but ‘That’s funny…’” The quip points to an media exposure across channels can’t be standards.
Many numbers mean little until we can important truth: if we treat the open question underestimated; advertising campaigns
compare them with a more familiar quantity. as more interesting than the neat answer, benefit from the multiplier effect, which Ensure the audience data used for trading
How big is the number compared with other we’re on the road to becoming wiser. comes from using a variety of media is based on metrics and methods that
things I might intuitively understand? How channels. The IPA Effectiveness Awards show comply with agreed and open industry
big is it compared with last year, or five years In the end, my postcard has 50-ish words that the average number of media channels standards.
ago, or 30? and six commandments. Simple enough, Source: Tim Harford 2018 used in a successful ad campaign continues Insist those standards are upheld by
I hope, for someone who is willing to make Tim Harford’s guide to to grow each year. independent auditing.
Finally, beware statistical significance. The an honest effort to evaluate – even briefly – statistics in a misleading Hold proprietary data sets to the same
simplest point to appreciate is that a number the statistical claims that appear in front of age 8th February. Data users are concerned, specifically when level of accountability as industry-owned
can be statistically significant while being them. That willingness, I fear, is what is most Used under licence from it comes to data collection and reporting trading currencies.
of no practical importance. Particularly in in question. the Financial Times. techniques. These concerns are fuelled Support the principles of the Joint Industry
the age of big data, it’s possible for an effect All Rights Reserved. by a number of specific issues around the Currency (JIC) model as the best-in-class
to clear the technical hurdle of statistical Broadcasters’ Audience provision of online and proprietary data sets; approach to providing objective and
significance while being tiny. Research Board Ltd is are these sufficiently objective for advertisers? comparable audience data and metrics.
solely responsible for Encourage all media owners and platforms
One study was able to demonstrate that providing this abridged Data transparency is challenging in an to engage with the Joint Industry
unborn children exposed to a heatwave while version of the original increasingly connected world where Currencies (JICs), not to the exclusion of
in the womb went on to earn less as adults. article and The Financial intellectual property is fiercely guarded, their own data sets, but to enhance their
The finding was statistically significant. But Tim Harford Times Limited does not speed to market is essential and data privacy credibility with advertisers.
the impact was trivial: $30 in lost income per Author, journalist accept any liability for is paramount. Media companies should
year. Just because a finding is statistically and broadcaster the accuracy or quality of understand the commercial benefits of The full report can be downloaded from
robust does not mean it matters. the abridged version. building a relationship with advertisers based


Measurement is a fascinating, oft- video, but the TV set was we need to think about the types of thing we Regulation may also thwart data sharing. Our ALL VIDEO
overlooked topic, which doesn’t get enough the preferred screen among measure and why. Mobile Consumer Survey indicated a lack of
attention. This matters as measurement is all age groups for watching understanding by consumers about the data
vital to the television industry, and to the long-form content, be it live, There are many hundreds of millions of they are sharing with services. Three quarters N E C E S S A R I LY
media sector more generally. catch-up or on-demand. In screens that can show video around the world, of UK adults claimed that they know their EQUAL, SO WE
2022, TV sets will still be in but each video format has different attributes. data are being used by companies and they NEED TO THINK
The TV advertising market is worth billions of 90% of homes and television A video watched on a smartphone screen will are aware of the risks. However, half said they
pounds in the UK, tens of billions of dollars in will continue to deliver the have a different impact to one seen on a 50- never share their email address online, yet
the US, and hundreds of billions globally. watercooler moments that inch screen. In turn, there is a question about they are on social networks. Two thirds said OF THING WE
advertisers crave. the size of the screen and the effectiveness they don’t share their phone numbers but MEASURE AND
This means that getting measurement of the advertising message. There is ample they are on WhatsApp. WHY
right for a digital age is critical: trusted Despite this, over the next research showing the differential impact of
measurement is fundamental to the industry, four years, the amount of content watched on screens of different sizes. The advent of GDPR this year will also have
but the job is far from done – video content traditional television viewing For me this says that they are of different a big impact on how we collect data. It will
remains in a state of metamorphosis, and is likely to fall. Research value and measuring the audiences needs to necessitate a different approach: greater
advertisers are arguably befuddled. from Thinkbox on video recognise these differences. consent, awareness and collaboration
consumption in 2016 showed around the data sets you are collecting
How is video consumption likely to look that 16-24-year-olds are We are unlikely to have a consistent and this will have an impact on audience
by 2022, and what will the imperatives for watching less traditional television than methodology for measuring everything by measurement and the data sets we can
measurement be then? 2022 is just four years the UK population as a whole, in favour of 2022, so we need to start making choices on expect to see in 2022.
away, so it’s tangible enough to influence consuming video content via other methods what to measure, from screen size and audio
industry decision-making today. such as SVOD and social media platforms. quality to video visibility and prominence and Imperatives for measurement
Our Mobile Consumer Survey indicates adjacent content. In conclusion, what are the imperatives for
Video consumption in 2022 that the behaviour of young people is measurement in 2022 for the industry as
Firstly, what will video consumption look replicated by all ages around five to eight Video measurement in 2022 a whole? It is important to recognise the
like in 2022? We anticipate that it will have years later. Measurement is likely to be as important in significance of measurement today and in
reached a new peak in terms of minutes of 2022 as it is today, if not more. There will be the medium term, and to understand what
viewing per person per day. This viewing will This means that while there will be more more data, richer data and more data sets advertisers expect and the digital platforms
be more atomised than ever, spread across video consumption in 2022, less than half of it than ever before, but these will be owned currently deliver. We must then evaluate the
an ever-widening array of devices. may be accounted for by live television. There by multiple parties, and they may not be consequences if television measurement fails
will be a huge increase in the range of SVOD inclined to share. to match what digital platforms offer.
One reason for this new peak will be better services available.
mobile networks, with improvements to By 2022 we will be able to collect more data Data sets are proliferating, but not all data
4G and the introduction of 5G, and better Broadcasters in the UK and worldwide are sets than ever. It’s technically possible that are equally important, and not all sets will Ed Shedd, Managing
smartphone screens, both of which will make currently looking at the success of SVOD the entirety of a purchase journey could be be available to access or link together. This Partner, North West
it even easier to watch video on the go. services and the reduced technology costs
IN 2022, TV SETS tracked from view to action to transaction lack of collective agreement will mean Europe TMT Practice
and deciding to follow suit and go direct to WILL STILL BE IN across multiple channels, devices and people will have to identify which data sets at Deloitte UK
However, while smartphones will be the consumers. Disney is launching an ESPN 90% OF HOMES networks, even if the purchase was weeks, and combinations are really important for
default for short-form content, TV sets SVOD service this year and a Disney branded AND TELEVISION months or even years after the first view. the success of their businesses, and then
will likely remain the dominant device for SVOD offering in 2019. We predict many The problem is that these data sets are in have collaborative conversations with
watching long-form content. content creators will similarly be diversifying WILL CONTINUE different silos. Data sets could be merged, their owners. Finally, we must establish
into SVOD, often as a complementary TO DELIVER THE but will the various companies that own the available technologies for delivering
In Deloitte’s Mobile Consumer Survey 2017, offering. W AT E R CO O L E R these data permit it? Unless we change the measurement and execute it. These
we asked interviewees in the UK which was conversation, there will be few collective data imperatives will hopefully stimulate us
their preferred device for certain activities. We must also consider that video is utterly
M O M E N TS T H AT sets, which will impinge on the effectiveness to have more collective conversations
Smartphones were the preferred device heterogeneous which in turn means that all ADVERTISERS of audience measurement and businesses and get much further in audience
for many activities, including short-form video impacts aren’t necessarily equal, so CRAVE that depends on it. measurement.

Left: ITV’s Love Island
© ITV plc

I’m hoping that there is a decent overlap So, how did this research process work on
between readers of BARB’s Viewing Report a show like Love Island? At the beginning,
and devotees of Love Island; I can imagine our network wanted a new reality show for
the Venn diagram. I love a good Venn ITV2, so we held brainstorms to stimulate
diagram. ideas. We pulled huge stacks of BARB data
together, mapping the gaps in the market 22% OF ALL
If you are still reading on and aren’t in in and sprinkling on qualitative insights from VIEWING TO EACH
my sub group don’t worry, I’ll explain. I’m our youth panel. When the decision was
about to get ‘extra’ over BARB. Which in Love made to commission a re-booted Love Island,

Island parlance means I’m pushing things, we tested the brand with our BARB-derived THROUGH A NON-
perhaps taking something too far, by being target audience and then fed this insight into TV DEVICE
over-the-top. the marketing campaign. We also made sure
that the press team had all the important those unique shows that has high device- We clearly have an exciting opportunity to
So, why am I getting ‘extra’? I sit on BARB’s facts so they could brief journalists quickly based online viewing. In the latest series, live take media research and consumer insight to
board, so I must declare an interest, but I when the ratings started to flood in. and on-demand viewing via devices added the next level, enhancing – and potentially
want to address two points. an average of 700k viewers per episode. even revolutionising – our understanding
We were able to do some of our most This means that 22% of all viewing to each of audiences and their interactions with
The first is to reveal just how fundamental effective work when the show was on air, episode was through a non-TV device. At messages and brands. For example,
BARB data are to a broadcaster/producer like combining minute-by-minute BARB data with present, we combine BARB data with our understanding the relationship between
ITV, using Love Island as a case study; and the insights from our viewer panel and online own ITV Hub statistics to find out how many media exposure and consumer behaviour
second is to illustrate how exciting the future ITV community, to help our commissioning people watched on all of our platforms. Soon via data overlays, described as the holy grail
is for those of us working in audience data and production teams tweak aspects of we will have Dovetail in place to make this of single source with large sample sizes, is
and insight, as we build brands on the back of the show, from format and pace to set and much easier. now a reality.
this strong currency. presentation. This really helped the show to
engage with and grow its audience. At the Online viewing through the ITV Hub is clearly We will continue to have BARB data at the
To explain the importance of BARB to a end of the series, we got together with all key very important to Love Island, and as we heart of any of these new opportunities. As
broadcaster, let’s start with the obvious. stakeholders to review the performance data match our registered user data with ITV Hub in the Love Island case study, the BARB data
You know that BARB data come in overnight; in tandem with qualitative insight in order to viewing we can get really granular detail on offer us the springboard to mix and match
what you may not appreciate is how that gauge viewer perceptions and expectations who is watching online. This is where things the appropriate data and insight to create
early morning email can change the whole for the next series. get really interesting, as we can start to something that amplifies the strengths of all
temperature of the company and the hyper-target audiences for advertisers and for datasets to create a whole bigger than the
workload on that day. You get hooked Naturally, we did lots of research to ensure marketing our new shows. sum of its parts.
on a rush of endorphins when a show we got the maximum value from the show
launches with a bang or grows week-on- and to prove to our clients the great benefits Audiences of all ages are not just embracing My hope is that this goes some way to
week; I won’t mention how it feels when they received from being associated with new platforms, they are demanding them; explaining why it feels odd to anyone working Neil Mortensen,
B A R B DATA a hit misses! the brand and advertising within it. And of the ITV Hub is on 30 platforms and we in a broadcaster to hear media experts Director of
OFFER US THE course, we wanted to make sure we had have over 21 million registered users of the talking about BARB not being relevant or Audiences, ITV
SPRINGBOARD TO Our audience insight process follows the the best information to help ITV’s service, as well as 10 million downloads of as important as it once was. I have been in
M I X A N D M ATC H full life of a programme, from development, international distribution team sell the companion apps for our biggest shows. many a meeting or conference when media
commission, production and broadcast format all over the world. Results from people say that they don’t believe in it. My
T H E A P P R O P R I AT E through to recommission, advertising impact BARB data formed the basis of this pitch Broadcasters now have a multitude response is that it isn’t a belief system, it’s
DATA A N D I N S I G H T and international distribution. The lifecycle and helped us to demonstrate how well the of opportunities to capture data from mathematics.
TO C R E AT E A of a programme doesn’t always follow a format had launched. interactions with their audience. ITV is
linear process like this, but we work on every becoming increasingly data-rich as it moves its If you don’t believe in maths, that’s up to
WHOLE BIGGER show at some point in its life, even if it is The other major part of the picture we deliver shows online, leverages social media, deploys you, but – to use a final bit of Love Island
THAN THE SUM OF simply measuring how many viewers came to is the overall programme performance for apps and other interactive solutions and lingo – you may find things a little ‘muggy’
I TS PA RTS it (which isn’t so simple!). this and every show. Love Island is one of explores advanced advertising technology. without it!


On-demand pre-broadcast Live On-demand post-broadcast


Episode 2

Episode 3

Episode 4

Episode 5
These days, viewing on-demand isn’t just
about catching up on programmes that have Episode 6
already been shown in the linear schedule.
0 100 200 300 400 500 600 700 800 900 1000
Increasingly, broadcasters are making the
whole series of a show available on-demand Individuals (000s)
on the day the first episode is transmitted
Source: BARB panel, 2017/18. Timeshift (excl Sky On Demand) and on-demand post-broadcast to 28 January 2018
– so people can get ahead with a series,
viewing episodes prior to their appearance
in the linear schedule. This is a practice
sometimes referred to as stacking.
BARB can now track these new patterns of
TV set viewing behaviour that combine on-
demand, live broadcast viewing and catch-up Total On-demand pre-broadcast
– perhaps predictive of the shape of viewing on-demand post-broadcast)
in an on-demand world. As percentage of adults (%)
16-24 16-24 16-24
The Tunnel: Vengeance is the third series of
25-34 25-34 25-34
the British-French crime drama The Tunnel. It
aired on Sky Atlantic in December 2017, with 35-44 35-44 35-44
all six episodes released on-demand 45-54 45-54 45-54
on the day of the broadcast debut of saw a complete shift across the life of the Above: Sky Atlantic's 55-64 55-64 55-64
episode one. series. Two thirds of viewing to the final The Tunnel: Vengeance
65+ 65+ 65+
© Sky UK Limited
two episodes was accounted for by people 0 5 10 15 20 25 30% 0 5 10 15 20 25 30% 0 5 10 15 20 25 30%
Fans of the first two series of The Tunnel watching those shows pre-broadcast.
were no doubt looking forward to the debut
episode of series three. Of course, they Drama series generally have a slight skew Gender
couldn’t watch episode one in advance, but (55%) to upscale ABC1 audiences but looking
Women Men Women Men Women Men
hardly anyone viewed it live, depending at the demographic profile of total TV set
on timeshift or on-demand to catch up viewing to The Tunnel: Vengeance, we see
afterwards instead. that over 70% of viewing was ABC1, with a
50/50 gender split.
And that tiny proportion of episode one Social group
viewed live was characteristic of the whole Indeed, the pronounced middle-aged C2DE C2DE
series, with even smaller percentages for the bulge would also be typical of traditional
later episodes, leading to a series aggregate TV viewing to drama series, but clearly
of only 3% viewed live. those viewers are making the most of new TRACK THESE
technologies, with more than half of on- N E W PA T T E R N S O F
That’s because the series was viewed almost demand pre-broadcast viewing to the series TV SET VIEWING
entirely on-demand, whether pre- or post- accounted for by the 45-64 age range.
broadcast (70% of total series viewing) or B E H AV I O U R T H AT
timeshifted via PVR (27%). We can also now examine the demographics COMBINE ON-
of online viewing on tablets and PCs because DEMAND, LIVE
But while the small proportion viewed we have installed software meters on our
live was consistent across the series, the panel members’ devices. This information
difference between viewing in advance can be combined with the census-level data VIEWING AND
of broadcast and catch-up afterwards we have for viewing on these devices. C ATC H - U P Source: BARB panel, 2017/18

Bake Off
Online viewing profiles* vs linear viewing profile ** © Mark Bourdillon / Love
Productions / Channel 4

Audience profiles for online TV viewing

Total PC Laptop Tablet Linear**

Adults 16+ Children 4-15

As percentage of adults (%)

To give an example, census data tell us higher amongst women, who accounted for
0 10 20 30% 0 10 20 30 40 50% 0 10 20 30% 0 10 20 30% 0 10 20%
that the 2017 final of The Great British almost 76% of total viewing minutes. And
Bake Off on Channel 4 achieved 311,000 while linear viewing to this show was skewed
average programme streams, but without strongly to upmarket viewers (60.8%), the
BARB panel data, we don’t know how many proportion of online TV viewing by those
Total PC Laptop Tablet Linear** people that was or who they were. BARB viewers (77.5%) reminds us that this typically
panel data tell us that this represented time-poor audience was determined not to
Women Men Women Men 397,000 people viewing that episode, miss out on the Bake Off final.
which – when combined with the panel’s
demographic profile of online TV viewing – Looking at the age and gender profile
can give us the demographic breakdown we of those using each device to watch
see in the chart. this particular show, we can see that a
disproportionate amount of viewing on
The Great British Bake Off has long been tablets is accounted for by younger viewers
credited with boosting the popularity of (16-24s represented 36.8% of tablet viewing
baking amongst younger audiences, and to Bake Off); whilst viewing on laptops was
ABC1 Adults
this was reflected with a healthy 27% of also high amongst 16-34s and children 4-15;
77.5% 78.2% 78.1% 76.0% 60.8% linear viewing to the final accounted for by and more than half of viewing on PCs was
young adults 16-34. Clearly, for lots of these by 25-34s – no doubt driven by workplace
younger viewers, linear viewing wasn’t an catch-up.
option as they were responsible more than
half (53.8%) of all online viewing on devices. The reporting of non-linear viewing on TV
sets like in The Tunnel: Vengeance example EXAMINING THE
And when we say young, it’s clear that is a core BARB panel capability. This is likely DEMOGRAPHICS OF
viewing this show online was a big hit for our to become increasingly important as more ONLINE VIEWING
youngest viewing segment of children aged broadcasters decide to make entire series
4-15, accounting for nearly 10% of minutes. available on-demand on the first day of linear IS POSSIBLE
With over 46% of its online viewing in the transmission. B E C A U S E PA N E L
25-44 segment, it looks clear that the final of MEMBERS ALLOW
Bake Off was a big family online viewing hit. Meanwhile, The Great British Bake Off final
Panel data from software meters, 2017 – currently unpublished demographic insights are a preview of the
There are perhaps no surprises in the gender type of learnings that will come from the next THEIR VIEWING ON
Source: BARB panel, 2017 skew – online device viewing was much stage of Project Dovetail. TA B L E TS A N D P CS


P rogramme Average audience
– individuals (000s)


1 Gun powder Se r i e s 1 Epi s ode 2 925.4

2 Top of the Lake Chi n a Gi rl Epi s ode 6 860.4

3 Top of the Lake Chi n a Gi rl Epi s ode 4 843.3

4 Top of the Lake Chi n a Gi rl Epi s ode 5 817.0

5 Motherl an d Se r i e s 1 Epi s ode 6 811.1

6 Top of the Lake Chi n a Gi rl Epi s ode 2 802.7

7 Top of the Lake Chi n a Gi rl Epi s ode 3 760.0

8 Motherl an d Se r i e s 1 Epi s ode 5 759.1
We’re all used to hearing industry chatter dominated by drama series. In top place is
9 Motherl an d Se r i e s 1 Epi s ode 4 677.2
about the decline of TV viewing; and it’s true Gunpowder, which premiered on BBC One
10 Motherl an d Se r i e s 1 Epi s ode 3 666.9
that traditional TV viewing declined from in the autumn, and was highly acclaimed by
11 Top of the Lake Chi n a Gi rl Epi s ode 1 523.8
212 minutes daily for the average person in critics and popular with audiences.
2016 to 203 minutes in 2017. Yet the total 12 Motherl an d Se r i e s 1 Epi s ode 2 518.4

amount of time spent in front of the TV set is Multiple episodes of Top of the Lake (series 13 Room on the Broom 489.4

only marginally down, dropping by just two two – China Girl) dominate the top of the 14 Bad Educati on Se r i e s 3 Epi s ode 4 462.4

minutes a day. chart, perhaps not surprising given its overall 15 Bad Educati on Se r i e s 3 Epi s ode 2 444.4

popularity and critical acclaim following the 16 Bad Educati on Se r i e s 3 Epi s ode 3 413.7
The reason for that apparent disconnect is series two debut in 2017. Box set viewing
17 Cl i que Se r i e s 1 Epi s ode 5 393.0
what’s called unmatched viewing, which has seems to be baked into this programme – the
18 Stacey Dool ey In vesti gates Se r i e s 1 Epi s ode 8 376.2
grown fast in recent years, accounting for first series launched at The Sundance Film
19 Bad Educati on Se r i e s 3 Epi s ode 5 371.0
14% of all TV set activity in 2016, growing to Festival 2013 with a seven-hour screening
20 Cl i que Se r i e s 1 Epi s ode 6 367.5
16% in 2017. of the whole series. Here in the top 50 we’re
seeing what happened when the whole series 21 Sti ck Man 360.2

BARB defines unmatched viewing as was released on-demand on the day of its 22 The Tun n el : Ven gean ce Se r i e s 3 Epi s ode 6 360.1

when the TV is on but the content wasn’t first broadcast. 23 The Gruf fal o 355.3

broadcast as part of a linear schedule in the 24 Stacey Dool ey In vesti gates Se r i e s 1 Epi s ode 8 353.0

last 28 days. This includes gaming, viewing This show illustrates a pattern over the 25 Bad Educati on Se r i e s 3 Epi s ode 1 351.8

to SVOD services and shows that haven’t lifetime of a series that has both a linear 26 The Tun n el : Ven gean ce Se r i e s 3 Epi s ode 4 349.7
been broadcast in the previous four weeks. broadcast and on-demand availability. The
27 The Gruf fal o's Chi l d 339.2
However, BARB can now identify the last later episodes accrue the bigger numbers, a
28 Quacks Se r i e s 1 Epi s ode 2 337.9
of these and take it out of unmatched, if consequence of viewers having more time
29 Cl i que Se r i e s 1 Epi s ode 3 336.8
broadcasters provide us with their non-linear to see the programme on-demand pre-
programme assets. broadcast and less patience to wait for the
30 Murdered for Bei n g Di f feren t 330.8

linear broadcast to find out how it ends. 31 Cl i que Se r i e s 1 Epi s ode 4 329.9

The table shows the top 50 most viewed 32 The Tun n el : Ven gean ce Se r i e s 3 Epi s ode 5 328.5

non-linear programmes of 2017. We can A similar pattern is also in evidence for older 33 Reg g i e Yates: Hi dden Austral i a Se r i e s 1 Epi s ode 1 322.1

begin to examine the patterns of behaviour series like Bad Education. Later episodes 34 Li ttl e Bri tai n Se r i e s 1 Epi s ode 3 305.7

when viewers access libraries of on-demand still capture the biggest views on-demand, 35 Quacks Se r i e s 1 Epi s ode 6 302.1
content, including box-sets. perhaps because they were missed last time.
36 Love an d Drug s on the Street : Gi rl s Sl eepi n g Roug h Se r i e s 1 Epi s ode 1 296.1

37 Quacks Se r i e s 1 Epi s ode 1 294.9
The table features multiple appearances The online-only BBC Three also delivers
38 Sex , Drug s & Murder: Li fe i n the Red Li g ht Zon e Se r i e s 1 Epi s ode 4 293.4
of classic children’s programmes like The some shows with on-demand audiences not
39 The Tun n el : Ven gean ce Se r i e s 3 Epi s ode 3 286.6
Gruffalo’s Child and Stick Man. These far behind those for shows that have also
programmes made the top 50 for 2017 after had a broadcast airing. Here we see bold, 40 Stacey Dool ey In vesti gates Se r i e s 1 Epi s ode 0 284.7

being made available alongside a range of hard-hitting, factual content that captured 41 I Shot My Paren ts 281.6

other BBC content via the BBC iPlayer over big audiences for one-off documentaries like 42 Quacks Se r i e s 1 Epi s ode 3 281.1

Christmas. It’s easy to picture the scene with Stacey Dooley investigates and I Shot My BARB CAN 43 Thi s Coun tr y Se r i e s 1 Epi s ode 1 280.9

little ones, parents and grandparents all Parents. 44 Peopl e Just Do Nothi n g Se r i e s 1 Epi s ode 1 269.9
enjoying these timeless classics. But the top REPORT VIEWING
45 W1 A Se r i e s 3 Epi s ode 6 269.8
50 helps to illustrate the characteristics of on- This evidence of shifting viewer behaviour TO S H O W S T H AT 46

Jami e John son Se r i e s 1 Epi s ode 3 267.3
demand, box-set viewing behaviour across supports BARB's development of techniques HAVEN’T BEEN
47 Bad Educati on Se r i e s 3 Epi s ode 6 267.1
the year too. that enable the comprehensive measurement
of viewing across all devices. We welcome
BROADCAST IN 48 W1 A Se r i e s 3 Epi s ode 2 265.9

As we’d expect from what we’ve long known working with broadcasters and rights owners THE PREVIOUS 49

Bi l l i on ai re Boy Se r i e s 1 Epi s ode 1 261.0
about timeshift behaviour, the rankings are on the reporting of non-linear viewing. FOUR WEEKS 50

Gan g sta Gran n y Se r i e s 1 Epi s ode 1 257.3
Source: BARB panel, non-linear viewing via TV sets to programmes 10 minutes+ in duration, 1 January - 31 December 2017
*Table includes content from rights owners that have provided their programme assets for measurement by BARB


Program m e B roa d ca s te r Average P rogramme C h anne l Average
P ro gramme Programme
S trea m s ( 000s ) Streams (000s)

1 Three Girls 611

S e ri e s 1 Ep i sod e 1 BBC 1 Love Island 24/ 07/ 2017 249

2 Love Island S e ri e s 3 Ep i sod e 2 9 ITV 541 2 I’m A Celebrity - Get Me Out Of Here! 19/ 11/ 2017 ITV 182

3 Doctor Foster S e ri e s 2 Ep i sod e 4 BBC 478 3 Election 2017 08/ 06/ 2017 BBC1 158

4 EastEnders 2 6 /1 2 /2 0 1 7 BBC 420 4 Love Island: The Reunion 30/ 07/ 2017 ITV2 124

5 The Apprentice S e ri e s 1 3 Ep i sod e 8 BBC 386 5 Live Pl: Liverpool v Everton 10/ 12/ 2017 Sky Sports ME 122

6 Blue Planet II S e ri e s 1 Ep i sod e 1 BBC 383 6 Love Island: Aftersun 25/ 06/ 2017 ITV2 121

7 The Replacement S e ri e s 1 Ep i sod e 3 BBC 344 7 One Love Manchester 04/ 06/ 2017 BBC1 111

8 Line Of Duty S e ri e s 4 Ep i sod e 6 BBC 344 8 World Cup 2018 Qualifier: England v Slovakia 04/ 09/ 2017 ITV 100

9 Rio Ferdinand: Being Mum And Dad BBC 315 9 England Friendlies: England v Brazil 14/ 11/ 2017 ITV 94

10 The Great British Bake -Off S e ri e s 1 E p is o d e 1 Channel 4 291 10 Doctor Foster 03/ 10/ 2017 BBC1 85

11 Liar Ser ie s 1 Ep i sod e 6 ITV 280 11 Match Of The Day 08/ 12/ 2017 BBC1 82

12 Louis Th eroux Dark States Ep i sod e 1 BBC 280 12 Weather 12/ 08/ 2017 BBC1 75

13 Peaky Blinders S e ri e s 4 Ep i sod e 5 BBC 277 13 Live Mexican GP: Race 29/ 10/ 2017 Sky Sport s ME 71

14 I’m A Celebrity... Get Me Out Of Here! S er ies 1 7 E p is o d e 2 ITV 270 14 Live Nissan Super Sunday: Chelsea v Arsenal 17/ 09/ 2017 Sky Sports 1 68

15 The X Factor S e ri e s 1 4 Ep i sod e 1 ITV 256 15 The Apprentice 04/ 10/ 2017 BBC1 68

16 Love Island: The Reunion S e ri e s 1 E p is o d e 1 ITV 251 16 Live British & Irish Lions Tour 08/ 07/ 2017 Sky Sports 1 66

17 Strike - The Cuckoo’s Calling S e ri es 1 E p is o d e 1 BBC 223 17 Live WCQ: Wales v ROI 09/ 10/ 2017 Sky Sports ME 64

18 Our Girl S e ri e s 5 Ep i sod e 4 BBC 223 18 Live Bristol City v Man Utd Carabao Cup 20/ 12/ 2017 Sky Sports ME 62

19 The Only Way Is Essex S e ri e s 2 1 Ep i s o d e 1 5 ITV 214 19 The BBC Election Debate 31/ 05/ 2017 BBC1 62

20 Trust Me S e ri e s 1 Ep i sod e 4 BBC 214 20 The Great British Bake -Off 31/ 10/ 2017 Channel 4 61

21 Made In Chelsea S e ri e s 1 3 Ep i sod e 1 1 Channel 4 207 21 BBC News 12/ 08/ 2017 BBC1 59

22 The Apprentice M e et The Cand i d ate s: S er ies 1 3 BBC 205 22 Match Of The Day 2 13/ 08/ 2017 BBC1 57

23 Call The Midwife S e ri e s 1 Ep i sod e 1 3 BBC 196 23 Ten O’Clock News 03/ 10/ 2017 BBC1 55

24 Mrs Brown’s Boys S e ri e s 1 1 Ep i sod e 1 BBC 194 24 Liar 16/ 10/ 2017 ITV 50

25 Made In Chelsea: Ibiza S e ri e s 1 Ep i s o d e 5 Channel 4 194 25 Blue Planet II 12/ 11/ 2017 BBC1 46

* Table only inc l ud e s si ngl e hi ghe st rate d e p i so d e p er p ro g ra m m e t it le a n d s p o r t in g e ven t *Tabl e on l y i n cl u de s s i n gl e h i gh e s t rate d e pi s ode pe r program m e t i t l e an d s por t i n g e ve n t

Viewing to personal computers, tablets and smartphones Viewing to personal computers, tablets and smartphones
Source: BARB TV Player Report 2017 - weekly data Source: BARB TV Player Report 2017 - weekly data


Live streaming On-demand

Device viewing
Live adds approximately

peaks due to
General Election
viewing to TV set
Billion minutes per week viewing

1.4bn Love Island

BARB’s TV Player Report is the UK’s as the UK general election, drive spikes in live finale
first joint-industry, audited measure of streaming. Six Nations I’m a Celebrity
viewing to online TV and reports on the Championship finale
consumption of TV content on PCs, tablets Because on-demand represents people starts
and mobile phones from data generated by having total control over what, when and v

software code embedded in TV player apps. where they watch, it’s not surprising that the
These are census data, not panel data – biggest programmes are the most popular,
providing a measure of total viewing across talked-about ones, across genres (see Top
devices by all users of the players, not just On-Demand Programmes and Top Live 1.0
by BARB panel members. Programmes, pages 20-21).

Viewing on devices has been much talked Drama (Three Girls, EastEnders, Doctor Foster,
about in our industry and is conspicuous etc) dominates the top on-demand rankings
because it’s often being done in public or on along with event TV (The Apprentice, I’m a 0.8
the move, leading to a belief that it might Celebrity and the Channel 4 debut of The
be higher than it actually is. In fact, the vast Great British Bake Off); with big numbers also
majority of viewing time is still happening on for one-off documentaries (Rio Ferdinand:
the TV set; our data show that total viewing Being Mum and Dad), natural history (the
time on devices adds approximately 1.3% to Blue Planet II debut) and entertainment
total TV set viewing levels. reality (Love Island, Made in Chelsea, The Only
Way is Essex). In fact, Love Island would have
Summer dip
Looking specifically at the share of online dominated the on-demand and live charts if in online viewing
TV viewing by devices, we can distinguish every episode had been treated separately. 0.4 reflects established
between BBC iPlayer viewing via smartphone seasonal trends in
apps, browser-based access and via the If the top on-demand programmes table viewing on
TV sets
iPlayer app on connected TVs. Our data show represents the content that people can’t bear
that 50% of all iPlayer viewing in 2017 was via to have missed, the top live programmes
connected TVs and this proportion appears to table tells us what they’ll make an absolute 0.2
be steadily growing. With connected TVs now priority for. Dominated by big football
available for 63% of individuals in TV homes games, rugby and Grand Prix racing, it’s again
(according to BARB’s Establishment Survey, not surprising to see Love Island and I’m a
Q4 2017), it makes sense that people will view Celebrity in the top 10. 0.0
online TV content on the biggest and best 1 Jan 5 Feb 5 Mar 2 Apr 7 May 4 Jun 2 Jul 6 Aug 3 Sep 1 Oct 5 Nov 3 Dec 31 Dec
available screen. We see that exceptional drama series like
Week ending 2017
Doctor Foster and Liar can command a
The iPlayer data on viewing behaviour are live streaming audience too – as well as a
consistent with data about which devices different kind of unmissable live drama in the
people prefer for certain activities. In form of the General Election exit poll, which
Deloitte’s Mobile Consumer Survey 2017 (see commanded big live-streaming numbers
Ed Shedd’s essay on pages 10-11) , the TV from 10pm on election night.
set was the preferred screen among all age TOTA L V I E W I N G
groups for watching long-form content. So online viewing seems to reflect two need- TIME ON
states that TV satisfies in us: the edge-of-the- DEVICES ADDS
The chart breaks down total viewing to seat live experience and the sit-back-and-
distinguish live streaming from on-demand. It relax personal viewing appointment. Let’s
shows that viewing on-demand accounts for now see what we can discover about when 1.3% TO TV SET See pages 20-21 for 2017’s top on-demand and live programmes
the vast majority, but particular events, such and how people are doing this online viewing. VIEWING LEVELS Source: BARB TV Player Report, 2017

Smartphone PC Tablet
Minutes viewed (billion) by clock hour

7.0bn Smartphone
6.0 Percentage Tablet
share across 46.0%
5.0 all audited TV
player apps
4.0 35.5%
BARB’s TV Player Report can tell us more us that while smartphones are always within 3.0
than total viewing levels for online TV – we reach, they are rarely first-choice device for
can also build a picture of when and how viewing TV content.
people are viewing.
So, device usage grows throughout the day
In the top chart, aggregated data for the and tablets are the device of choice; but
whole of 2017 show us the pattern of viewing has that been changing over time? In short,
throughout the day on PCs, tablets and not really. The TV Player Report has been 0
06 08 10 12 14 16 18 20 22 24 02 04
smartphones. Broadly speaking, it shows an gathering data since 2015 and in the third Time of day
evening viewing peak, just like TV – although chart we show the trend for the past six
the viewing peak for devices (8-11pm) is later quarters.
Since Q3 2016 (and indeed consistently
Looking in detail at share of viewing by device since 2015), tablets have been the nation’s
at different times of the day, we see a picture favourite device for viewing TV content 60%
that’s consistent overall (tablets ranking first, away from the TV set – taking around 45%
PC second and smartphones in third place), of viewing time on non-TV devices. This Tablet
but with some interesting peaks and troughs. first-choice position for tablets looks to be 40
pretty unassailable, as viewing on PCs, the
Tablets peak in the morning – and it’s clear next most popular device, has declined 4.3 PC
that tablet viewing dominates breakfast time percentage points year-on-year from Q4 2016 20 Smartphone
and the morning commute, peaking at 7am to Q4 2017.
with 58% of device viewing, and dominating
all the clock hours from 6-9am with more In contrast to this, viewing via smartphones
than 50% share. It’s tempting to put this has been on the increase – climbing more 06 08 10 12 14 16 18 20 22 24 02 04
down solely to viewing while commuting, than three percentage points year-on-year Time of day
but from what we also know about the - most likely due to a combination of the
popularity of tablets amongst children, ubiquity of smartphones, the trend towards
perhaps there is also an element of tablets larger handset screens and cheaper data SHARE OF ONLINE TV VIEWING OVER TIME BY DEVICE
as breakfast-time childminders. packages.
Smartphone PC Tablet
PCs get their highest share of device viewing Nevertheless, the ranking of devices seems to
in the middle of the day between 12-3pm, reflect what we’d expect by way of the quality 50%
no doubt driven by workplace viewing in the of the television viewing experience – larger
lunch hour (but perhaps not just in the lunch screen sizes winning out over smaller and 40
hour!), peaking with 40% share at 2pm. By PCs losing out to the more personal devices
contrast, share of viewing on tablets slumps of tablets and smartphones. 30
at lunchtime (from 12-2pm) – with a share
of 41% at 1pm (just fighting off PC to retain But overall, it’s clear that we love watching 20
top place). In fact, in the second chart, we TV around the clock on whichever device we
see that the share of tablet viewing and PC can get our hands on; and, if we can, having DEVICE
viewing are inversely related to one another prime time all the time. Let’s take a look next USAGE GROWS
over the day. at who’s doing this online viewing. THROUGHOUT THE 0

Meanwhile, smartphones get their highest

DAY A N D TA B L E TS Q3 2016 Q4 2016 Q1 2017 Q2 2017 Q3 2017 Q4 2017

share (23%) at 2am and throughout the early ARE THE DEVICE
hours of the morning from 1-5am – reminding OF CHOICE Source: BARB TV Player Report, 2017


On-demand Live streaming Total



BARB’s TV Player Report has been providing inclined to view live, while viewing on-
device-based census data about online demand satisfies the desire among younger
viewing since 2015 – from the UK population people to view when (and where) they want. 20
at large. From 2018, BARB will tell us
exactly who is doing this online TV viewing. We can see a gender difference too between
on-demand viewing and live streaming, 10
This is the outcome of Project Dovetail, which with men much more likely to view live –
links the device-based census data to BARB’s largely driven by their interest in sports
people-based panel data to provide the UK’s programming – and women more likely to 0
only authoritative source of the demographic catch up with their chosen programming Women Men Age 4-15 16-34 35-54 55+
profile of online TV viewing. While BARB on-demand. This is in keeping with long-
hasn’t yet been reporting this information, established genre differences between live
we have been gathering it for some time and and timeshifted viewing.
so we can share some findings from the BARB
panel on just who it is that is viewing on PCs Patterns of online TV viewing over the day
and tablets (we don’t yet record viewing via bear a strong resemblance to the shape of TIME OF DAY VIEWING BY AGE – PC AND TABLET VIEWING
smartphones for the BARB panel). These TV set viewing generally – dominated by Average viewing (000s)
data are taken from the whole of 2017. evening viewing, although a little later than
traditional TV. The peak in online TV viewing
The top chart shows the gender and age for adult audiences is between 9-11pm, which 50
profiles of online TV viewing. Looking at builds earlier (from around 8pm) for older
total online TV viewing, we can see that it audiences (55+), but is concentrated into a
skews to women more than men (perhaps shorter, earlier peak (9-10pm) for younger 40
challenging the convention that men, as adults (16-34). This 9-10pm spike in young
TV’s lighter viewers, might have been more adult viewing (followed by a sharp drop)
likely to turn to online viewing), and to older is accounted for by live streaming to Love 30
audiences too. However, when compared Island (in June and July) and I’m a Celebrity
with traditional TV set viewing profiles, the (in November and December). With fear of
online picture is more as we might expect. missing out (and of spoilers on social media), 20
these reality series are clearly something you
Children’s share of online viewing by device have to see live. Adults 55+
(8.4%) leads their share of traditional TV set 10 Adults 35-54
viewing (6.5%), while 16-34s take a far higher These articles on online TV viewing have
share of device viewing at 24%, compared given us a glimpse into the new insights A D U LT S I N T H E Adults 16-34
to 15% of that via a TV set. Figures match that Project Dovetail will bring in 2018. 0 Children 4-15
more closely for 35-54s (taking 33% of device Soon we’ll be able to compare - across
06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 24:00 02:00 04:00 06:00
viewing and 28% of TV set viewing), while on-demand, live-streaming and linear - the WORKING LIVES
Time of day
the over 55s take 51% of TV set viewing, viewer profiles of schedule stalwarts like APPEAR TO BE THE
compared to 35% of device viewing. Adults in EastEnders and Coronation Street that GROUP WHOSE
the prime of their working lives, possibly pre- dominate the rankings in all environments,
parental responsibility, and most likely to be and see who’s viewing what. ONLINE VIEWING
out and about, appear to be the group whose BEHAVIOUR IS
online viewing behaviour is diverging from DIVERGING FROM
their TV set behaviour most sharply.
But age does seem to drive different viewing BEHAVIOUR MOST
patterns – we see that older people are more S H A R P LY Source: BARB panel – PCs & tablets, 2017

Households with at least one subscription (millions)
Year-on-year change

12 Q4 2017 v Q4 2016

10 +23% Any SVOD*

8 +25% Netflix

In Bring on the box sets (pages 18-19), we We can also compare the SVOD services 1

highlighted the fast-growing area of activity by their total annual subscription growth. In the UK, television is still king. We are investing in ever larger TV sets; more
than half of all UK households have a TV at least 40 inches wide and 16% of
homes have a set of at least 50 inches. We’re getting a lot of use out of them too;
4 +41% Amazon
called unmatched viewing, which includes Netflix, still the dominant SVOD player in the THE SVOD
in 2016, daily total TV set viewing by individuals was four hours 12 minutes on
average. But how and what we are choosing to watch is changing.

viewing more than 28 days after linear UK, grew at the slowest rate of the three main GROWTH IN SVOD
Average minutes, BARB panel
300 Unmatched

broadcast and non-broadcast viewing. services, with a 25% annual growth rate,

8–28 day
January 2018 200 viewing

Viewing to subscription video-on-demand but added the most subscribers in the year, +40% Now TV
150 Live & 7-day
100 viewing


(SVOD) services like Netflix and Amazon is rising by 1.6 million to a total of 8.2 million 0
2015 2016

part of unmatched viewing and is likely a households. Amazon and Now TV were In a world of fragmentation, traditional TV viewing is declining (chart 1). In
2015, the average person watched 216 minutes daily of traditional consolidated 0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
TV (live and viewed within seven days of broadcast). In 2016, this fell to 212

significant contributor to this growth. neck-and-neck in their annual subscription

minutes, a 2% decrease, although the total amount of time spent in front of the
TV is unchanged. The reason for this apparent disconnect is that a fast-growing
area of activity is unmatched viewing.

growth. Amazon grew 41%, adding 1.3 million 2 UNMATCHED VIEWING

Unmatched viewing as a % of total TV set usage, BARB panel
2014 2015 2016 2017
Unmatched viewing accounted for 16.4% subscribers to reach a subscription base of


15 Source: BARB Establishment Survey

of all TV set activity in 2017 and continues 4.3 million households, while Now TV grew by 10

*Any SVOD = any of Netflix, Amazon or Now TV
to grow. For example, during the spell of 40%, adding 0.4 million subscribers to reach 0

snowy weather this March, total TV viewing a base of 1.5 million households. Unmatched viewing accounted for around 14% of all TV set activity in 2016,
and this figure has reached almost 19% in recent months (chart 2). While a

(live and seven-day catch-up) increased by

significant proportion of this can be attributed to games-playing on a games
console, the rest ranges from archive playback of PVR recordings over 28 days
old, to DVDs and, most importantly, to watching programming via subscription
video-on-demand (SVOD) services.

8.4% year-on-year in the week ending 4th Looking at SVOD access by age group, 55% SVOD ACCESS BY AGE GROUP
March 2018. However, total screen time of all children in the UK have access to one
increased by 14.8% during this time. This led of the SVOD services. For young adults aged Any SVOD* Netflix Amazon Now TV
to unmatched content accounting for 19.5% 16-24, 62% have access to one of the SVOD Above: BARB SVOD Report
of TV set usage. For under 25s, unmatched services, while the figure is 56% and 52% for
content accounted for nearly 43% of TV set 25-34s and 35-44s respectively. Deloitte’s Children
usage in that week. Mobile Consumer Survey indicates that the
behaviour of young people is replicated by all
While we aren’t yet able to measure actual ages five to eight years later (see Ed Shedd’s 16-24
viewing to SVOD services without the guest article, pages 10-11), so SVOD access
cooperation of the service providers, we can by older groups is likely to increase over this
get a sense of the popularity of these services time too. 25-34
from our Establishment Survey, which has
been asking questions about SVOD service With SVOD services in 36% of UK households
take-up since 2014. and continuing to grow, it’s clear that SVOD 35-44
isn’t a new entrant or optional; it is a staple
The SVOD market is enjoying a sustained service. Far from being niche, SVOD services
period of growth in the UK; the number of are now an established part of the television 45-54
households that subscribe to at least one ecosystem.
SVOD service surpassed 10 million in Q4 2017,
a 23% increase on Q4 2016. When you take To read BARB’s recent SVOD report, featuring 55-64
into account homes that have more than one further analysis of SVOD access in the UK,
service, we can project that there are at least visit: WITH SVOD 65-74
14 million subscriptions to SVOD services. svod-report-charting-the-growth-in-svod- SERVICES IN
This is in addition to any pay-TV packages services-across-the-uk/
to which homes are committed; indeed, our 36% OF UK
data show that there is significant overlap To see the latest data on SVOD household HOUSEHOLDS AND 75+
between households that choose to pay subscriptions, visit: CONTINUING TO
for television and those that also choose tv-landscape-reports/tracker-svod/
to pay for SVOD services. Many households
G R OW, I T ’ S C L E A R 0% 10% 20% 30% 40% 50% 60% 70%

subscribe to both to get access to all of the T H AT S VO D I S A Source: BARB Establishment Survey, Q4 2017
content they want. STA P L E S E RV I C E *Any SVOD = any of Netflix, Amazon or Now TV

You can also now look at viewing levels over
time by device. This shows a spike in viewing
on tablets and PCs during the week of last
year’s general election, 5th – 11th June. If

we were to look at these data day-by-day,
no doubt we’d find this peak on Thursday
evening/Friday morning, driven by night
owls keeping up with the twists and turns of
another unpredictable night of politics.

If this has whetted your appetite for more

online viewing data, you can explore the TV
Player Report on our website, visit http://

How we measure Donaldson and illustrator Axel Scheffler’s
viewing to non-linear picture books, topped the chart with report-2/
programmes average audiences of 328,000 and 295,000

In the cold and dark days respectively. The top ten also includes Understanding how
of winter, it’s tempting to three entries for The Tunnel: Vengeance, people watch on PCs
forgo an evening out in while the top 30 is dominated by dramas on and tablets
favour of curling up on the sofa and getting iPlayer, from the sole newly-released title, Previously, we looked at
stuck into a box set. In December, the BBC Feud: Bette and Joan, through to previously how we collect the online
brought back a host of programming to broadcast programming such as Sherlock, TV viewing census data
iPlayer for the holiday period, ranging from Line of Duty and Happy Valley. that go into our TV Player Report. These data
classic dramas and comedies through to represent total viewing across PCs, tablets
natural history and family programming, One caveat to note is that if any of the titles and smartphones, but they don’t tell us
while Sky offers hundreds of drama box sets were available on another service at the about the people doing the viewing. How
to its customers. same time, we cannot distinguish which do we know who they are and how many
service viewing took place on; for example, people are watching? To find this out, we
At BARB, we are able to measure non- Line of Duty was available to watch on Netflix must go to BARB’s nationally-representative
linear viewing on a TV set like this in the as well as on iPlayer in this period. panel of homes.
same way that we measure traditional
broadcast television viewing, by using an But whether it’s adults catching up on Many European countries use several panels
audio-matching process. The software a recent drama or the kids enjoying an Above: BBC1’s McMafia – the a TV player app, be it live or on-demand, to monitor viewing on different devices. In
meters attached to the TV sets in BARB animated film, we can track which non-linear most-watched on-demand TV the embedded software creates viewing contrast, BARB’s strategy is to have a single-
programme in January 2018
panel households take an audio sample of programme is being viewed, and who’s doing statements detailing what has been watched, source for how people watch on different
© BBC Worldwide
the programme, which is then turned into a the watching. and, to the second, for how long. devices. This has the advantage of generating

digital fingerprint and matched to a reference better inputs, as BARB meets the need for
library of programmes. How we collect Kantar Media collect these data for BARB deduplicated reporting of programme and
online TV viewing so that we can publish data on the number commercial audiences across multiple
Normally, BARB builds its reference library data of PC, tablet and smartphone devices that screens. We also have to consider that
of programmes by monitoring the linear Did you know that if you, are being used to watch programmes, both combining data from multiple panels is likely
schedules. In the case of these non-linear or anyone in the UK, on-demand and live streamed. Importantly, to result in less precise demographic profiles,
programmes, we rely on the rights owners watch television through these are census-based figures; they which are key for our customers.
providing us with a copy of the programme a broadcaster’s online TV player app, we represent total viewing levels for participating
assets before they are made available can capture these viewing data? BARB’s TV online platforms on these devices. This is why the majority of the 5,100 homes
through an on-demand player app. The BBC Player Report is the UK’s only fully-audited, on our panel have software meters attached
has been working with BARB to measure joint industry measure of online TV viewing The TV Player Report data can be viewed to their PCs and tablets which track their
programmes in this way for nearly two years, on PCs, tablets and smartphones. It currently in a number of ways. You can compare the online TV viewing; panel members must
while Sky has recently started to use this reports total viewing time on PCs, tablets aggregate viewing time or live streaming register who, and how many people, are in
technique to understand how people are and smartphones, and also programme-level WE MEASURE by TV player/channel or broadcaster group, front of the device.
watching its box sets, providing programme audiences for online platforms such as All 4, NON-LINEAR examine these viewing levels over time or
assets for its crime thriller The Tunnel: BBC iPlayer, ITV Hub, My5, SkyGo and VIEWING ON A look at the top 100 programmes across all The panel data provide valuable
Vengeance over Christmas. UKTV Play. platforms or by TV player or broadcaster. demographic context for the TV Player
TV SET IN THE So, you can find out that the most-watched Report’s census data. For example, we can
Looking at the most watched non-linear How does this work? We generate online TV S A M E W AY T H AT live programme in January was the Premier look at the age profile for total viewing time
programmes for which the rights owners had viewing data from software code that has WE MEASURE League clash between Liverpool and to BARB-measured TV player apps via PCs
provided assets over the festive period, half been added to 40 different online platforms; Manchester City on Sky Go, and top of and tablets in 2017. This shows that for all
of the top ten were children’s titles on BBC these have been audited by ABC to ensure the
BROADCAST the on-demand chart for the month was the viewing via computer devices, around a
iPlayer. Room on the Broom and Stick Man, data meet our standards. Whenever someone TELEVISION first episode of crime drama McMafia on quarter came from 16-34s – who only account
the animated adaptations of writer Julia in the UK watches a programme through VIEWING BBC iPlayer. for 15% of all TV set viewing. For All4, this

figure was even higher, at 41% of all impacts are not attributed to advertisers owners to deliver duration-based information the net number or percentage of people who
viewing, highlighting the demographic whose ads have been swapped out for the on viewing sessions. have seen a specified amount of a particular
differences between the online players. dynamically served ads. piece of content; for television, this might
So how do we approach viewability in the be a programme, channel or advertising
We don’t yet track panel homes’ viewing Our measurement system gives all BARB context of these inputs? campaign. The amount of viewing that an
on smartphones. Smartphone users are customers the ability to see the extent to individual must have done in order to be
selective about which apps they download, which commercial impacts have moved into The first point is that our online TV metrics counted as having been reached varies by
so they may not wish to have software this new commercial format. For example, are duration-based. Our metric reports medium. The BARB definition is for this to be
meters installed on their phones to monitor our data show that Sky AdSmart impacts the average number of streams for both at least three consecutive minutes.
their viewing in this way. We believe that increased by 92% year-on-year, based on programmes and commercials; this takes
the risk of panel homes deciding to leave all individuals for the calendar month of into account the total amount of time that a This means that the reach figure for a
due to the perceived intrusiveness of a January in 2017 and 2018. In January 2018, programme or commercial has been seen and television programme is often much higher
smartphone measurement system is not Sky AdSmart ads accounted for 1.15% of all is analogous to established TV metrics such than the average audience figure, but arguably
matched by the reward in collecting these impacts sold by Sky. as average audience and TVRs. These metrics less meaningful. Reach provides an indication
data, as smartphone viewing makes up a tiny have been ratified by JICWEBS, the industry- of the number of people that watched at least
percentage of total television viewing. PC, We report the commercial impacts for owned body responsible for the development three minutes of a show – but they may have
tablet and smartphone viewing as a whole dynamically served ads at an aggregate level; of online measurement standards. not liked the show and so switched off, or they
add just under 1.5% to TV set viewing, and we don’t provide a verification service for may have just caught three minutes of the
we know that less than one-fifth of this is on individual Sky AdSmart campaigns. We are The next point is we only collect census-level middle of a programme while channel-surfing.
a smartphone. also in conversations with other platforms streaming data from tablets and smartphones By contrast, average audience represents the
and broadcasters with the objective of if the app is fully viewable on the screen. In total amount of time viewed, which is a more
We are exploring the use of router meters as a delivering equivalent reporting for their 2017 these devices accounted for 64.5% of reliable indicator of popularity; after all, if you
method of monitoring a household’s viewing dynamically served advertising propositions. time spent watching online TV, excluding the enjoy a show you are likely to watch more of
TV set; all of this is fully viewable.

across all devices, including smartphones. it. We can see a number of examples of this if
Our approach to

we look at the reach compared to the average
How we report viewability There is a potential issue when online TV is audience for various programmes that aired in
dynamically served It’s natural that viewed on PCs; last year this was 35.5% of January/February 2018.
advertising advertisers would only online TV viewing on devices other than the
BARB reports viewing to want to pay for an ad that TV set. An issue can arise when someone It is logical that average audience figures
commercials, known as has had the opportunity to interacts with content in multiple browser are considerably different to reach figures
commercial impacts, in be seen by a person. This is why viewability is windows or opens another application over for certain programmes. Match of the Day
the same way that we report viewing to an essential part of any media measurement the top of their web browser. has a higher reach than average audience,
programming. These data are relied upon system. as fans may tune in just to catch their team’s
by advertisers and agencies looking to Commercial broadcasters are working with game. By contrast, a film like Batman Begins
assess the return on investment of their At BARB, we consider a television programme industry-ratified vendors of viewability on Channel 5 has a slightly higher average
advertising campaigns. But how do we report or commercial to have been viewed on a TV products, that can help to identify when PC- audience, as viewers are engaged by the
dynamically served advertising? set when it is watched at normal speed, and based viewing of television programmes and narrative and so likely to watch the whole
at least one panel member is registered as ads is in the top window. BARB will work with broadcast. The ITV weather has a high reach
Dynamically served advertising is the ability being in the room. We deliver duration-based commercial broadcasters with the intention thanks to inheriting an audience from the
to swap one ad in a television programme metrics because our television ratings (TVRs) of building these viewability measures into news, but a lower average audience, perhaps
for another, in order to replace an ad seen are based on a technique that estimates the our reported viewing figures for online TV because viewers start switching off when they
by linear television viewers with one that is average number of people present across the commercials. remember that forecasts are unreliable! Why
targeted to a specific audience. BARB was the duration of the ad. the average audience for The Graham Norton
first television industry currency in the world We are working on producing multiple screen Show is so much lower than the reach is a
to deliver a measurement of dynamically The principles in this approach are important viewing data for commercials using the little harder to unpick, but given that 28% of
served advertising; we have been delivering a as we start to report viewing to online TV Dovetail Fusion methodology and will have viewing to this episode was timeshifted, it may
solution for Sky Ad Smart since 2013. commercials. This is because the industry is an update on this later in the year. be that viewers are fast-forwarding though
the show to get to their favourite guest, be it

asking us to apply these conventions, rather
How does it work? Working with Sky, the than using viewability standards based on
BARB WAS THE Why duration- Jamie Dornan or Dame Helen Mirren.
advertiser chooses the target audience the video content being partially in view for a FIRST TELEVISION based metrics are
for their dynamically served ad, based on minimum number of seconds. AUDIENCE important Nonetheless, advertisers come to television to
factors such as location, demography and CURRENCY IN Previously we explained build reach; television is widely acknowledged
lifestyle. Sky then flags when an ad that was Our reporting of online TV viewing relies on why duration-based to be one of the most effective media for
due to air in the linear broadcast is replaced two data sources. Firstly, we collect data from THE WORLD metrics are key to BARB’s addressing a large number of people. That D U RAT I O N - B A S E D
by a dynamically served ad in households software meters that are installed on tablets TO DELIVER A approach to viewability. However, duration- said, the frequency of seeing an ad is also METRICS ARE A
matching the advertiser’s target audience. and PCs in our nationally-representative MEASUREMENT based metrics are important for other important for brand-building and this is MORE RELIABLE
This substitution is picked up by the meter panel of UK homes: this provides evidence reasons. where duration-based commercial impacts
attached to the TV sets in BARB panel homes, of who is in front of the screen. Secondly,
O F D Y N A M I C A L LY- are used alongside reach metrics. And when it
so we know an ad has been switched. This we collect census-level streaming data from SERVED The chief alternative to duration-based comes to determining the cost of a campaign, POPULARITY AND
means we can ensure the commercial viewing apps that have been enabled by their ADVERTISING metrics is the reach measurement. Reach is the latter naturally takes priority. ENGAGEMENT


15% viewing in 2017

Summer drops
Every year in The Viewing Report, BARB examines UK television upviewing
0.5 percentage points
in 2017

viewing behaviour in the previous year from three key angles: 10 on 2016 but the lowest
Summer drops
5 increase this
Timeshift – how much timeshifted television did UK viewers decade
watch? When and why did timeshifted viewing peak and dip? up 0.5 percentage points
Which genre was the most timeshifted? And which 0
demographic and region were responsible for the most 2010 2011 2012 2013 2014 2015 2016 2017 on2018
2016 but the lowest
timeshifted viewing? 5 increase this
 hare by genre – How did the genres rank by share of total TV
S We continue to watch more timeshifted TV. Average 7-day tournament or Olympic Games this dip was smaller than in
viewing? Which genres saw significant increases or decreases timeshift viewing in 2017 was 14.3%, an increase of 0.5 previous years. With timeshifting more stable across the year, it
2010 2011 2012 2013 2014 2015 2016 2017 2018
in their share, and why? percentage points on 2016 – but this was the lowest increase is possible that audiences are turning their attention elsewhere.
this decade. We know that unmatched viewing is growing and can surmise
Share by broadcaster – How did the major broadcaster that SVOD services account for a large part of that. Netflix,
groups rank by share of audience? Which of their channels As usual, the summer months saw a dip in timeshifting, Amazon and Now TV may be impacting timeshifted as well as
had the highest share? TIMESHIFTING BY... of a major international football
although with the absence live viewing.

Over the next three pages, we’ll find out the answers to these Genre Demographics Region
questions for 2017 viewing. Drama All individuals South & South East
Arts Children 4-15 London
For further updates on viewing trends, check out the Trendspotting TIMESHIFTING BY...
Education East of England
tab on the BARB website, which features quarterly trackers of metrics Films Genre Adults 16+ ABC1 Demographics West Region
such as device access, SVOD households and UK households by TV Documentaries Adults 16+ C2DE UK average
platform. Entertainment
Drama AllAdults 16-24
individuals Midlands
South & South East
Arts Adults 25-34
Children 4-15 South West
Education Adults 35-44 North
East of East
Films Adults
Adults 16+ 45-54
ABC1 Yorkshire
Documentaries AdultsAdults 55+
16+ C2DE Wales
UK average
Entertainment Adults 16-24 North West
Music Men
Adults 16+
25-34 Scotland
South West
Children’s Women
Adults 16+
35-44 Border
North East
Party Political Broadcast
Hobbies/Leisure Houseperson with45-54
Adults Kids Ulster
Sport Adults 55+ Wales
Religious 0 5 10 15 20 25 30% 0 5 10 15 20 25% North West0 5 10 15 20%
Men 16+ Scotland
News/Weather Women 16+ Border
Party Political Broadcast Houseperson with Kids Ulster
Source: BARB

0 5 10 15 20 25 30% 0 5 10 15 20 25% 0 5 10 15 20%

Source: BARB

In 2017, drama retained its place as the most timeshifted genre. viewing, likely a reflection of time-poor professionals with
Adults 25-34 and the South & South East and London continued young kids and long commutes using timeshift to catch up on
to be the demographic and regions doing the most timeshift their favourite shows.

Source: BARB


Trend in % share of audience, 2008–2017 % share of audience, 2017
Percentage share of audience, 2017
In 2017, four unlikely characters have come
Entertainment 18.6%
together to influence our viewing habits:
President Donald Trump, David Davis MP, BBC One 21.8% Drama 1.2%
Trend in %Drama
share of audience, 2008–2017 16.3%
Syrian President Bashar al-Assad and
(series+soaps+single plays)
Prue Leith. They are all central characters 31.6 BBC Two 5.8% 5.4 Dave 1.1%

Percentage share of audience, 2017
14.1% in helping the factual television genres % Cbeebies 1.3% % Yesterday 0.6%

Entertainment 18.6% Others 2.6% Others 2.4%

(documentaries, news/weather and current
News/weather 10.5% affairs) reach a 29.3% share of viewing – more
Drama 16.3% than a four-point increase since 2014.
(series+soaps+single plays)
Films 8.5%
(cinema+tv+other) 14.1% Sport’s share of viewing fell, as expected in
Hobbies/leisure 8.0% an odd-numbered year, while the children’s
News/weather 10.5% television market continues to struggle, with ITV 15.6% Sony Movie Channel 0.5%
the genre dropping another 0.4% to follow
8.5% the 0.5% fall in 2016. As children’s unmatched 21.7 ITV2 2.3%
2.5 True Entertainment 0.4%
Movies 4 Men
viewing grows (reaching a daily average of % ITV3 2.0%
% 0.4%

53 minutes in 2017 from 43 minutes in 2016) Others 1.9% Others 1.3%

Hobbies/leisure 8.0%
we may be witnessing the impact of SVOD
Children's 4.1% services, such as Netflix, exerting downward
Sport 7.1% pressure on viewing to linear children’s
Music 0.8% programming. The FIFA World Cup will
4.6% undoubtedly see sport bounce back in 2018,
Arts 0.2% but another genre showing strong growth
Channel 4 5.5% Quest 0.7%
Children's 4.1% in 2017 was entertainment. Strictly Come
Religious 0.1% Dancing, I’m a Celebrity and a plethora of 10.2 Film4
2.3 Quest Red

Music 0.8% reality entertainment shows have driven the % 1.4%

% 0.3%

Education 0.0% genre forward in 2017. Others 1.5% Others 1.1%

Arts 0.2%

Religious 0.1%
Educationcover a wide
Documentaries 0.0%spectrum of interests, and attract significant audiences*
Channel 5 4.2% CBS Reality 0.6%

Blue Planet
Documentaries II reached
were Blue PlanettoII BBC
the fastest growing factual genre, skewed older,
One’s Blue while
Planet II. Below we take a look at who these 8.4 5USA 1.0%
1.4 Horror Channel 0.4%

appealed more strongly to women
increasing their share by 1.1 percentage points to reach 14.1% viewers were compared to the next biggest show broadly defined
% 5Star 0.6%
% CBS Action 0.4%
the highest level thisSTATS
decade. This was in no small part down as a documentary
Viewer index (against total universe)
– Channel 4’s The Great British Bake Off. Others 2.7% Others 0.1%
Documentaries cover a wide spectrum of interests, and attract significant audiences*
Blue Planet II reached Blue Planet II skewed older, while

appealed more strongly to women
reached 10

Viewer index (against total universe)
Sky 1 0.9% Universal 0.4%
Pick SyFy 0.2%
20 8.2 Sky Sports
0.9 Movies 24 0.1%
reached -30
% Main Event
% Others 0.2%

* All individuals aged 4+ that viewed for
Children Adults Adults Adults Men Women
at least three consecutive minutes 16-34 35-54 55+ 16+ 16+

-20 Looking at the top ten broadcaster groups by share of viewing in points. The main change sees NBCUniversal International
-30 2017, the top nine remain unchanged on 2016, with Viacom and Networks claim the tenth spot in the share ranking, with 0.9%,
* All individuals aged 4+ that viewed for Children Adults Adults Adults Men Women Sky swapping places as Viacom moved up to fourth place. The displacing Turner which held tenth place in 2016.
at least three consecutive minutes 16-34 35-54 55+ 16+ 16+ two remain very close in share, separated by just 0.2 percentage

Source: BARB

Where applicable, channels include HD and +1 variants.

Channel share within family may not add up due to rounding.
Source: BARB Source: BARB

BARB runs two free training sessions for sessions help to explain the potential of
subscribers at its offices in central London, the data, including how new developments
to help the media and advertising industry will shape the future of the BARB service.
better understand how we deliver a trusted It focuses on the following topics and
audience measurement currency. questions:

Boot Camp is designed to support the What demographics do we collect from our
emerging generation of media agency, panel?
broadcaster and advertiser talent. It gives an How are the BARB data structured and how
introduction to BARB and what we do: how can they help you analyse TV viewing?
the BARB panel is formed, how we measure Tips on segmenting and creating dynamic
viewing from panel households and how audiences.
we collect census data from TV player apps. How the technology behind the panel is
It also offers an insight into our plans for adapting to keep up with changing viewing
the future. This session is for those who are habits.
new to the industry or want a refresher on Using the BARB Establishment Survey to
the basics of how we measure and report understand your potential audience and
audiences. size.
How are SVOD services growing and
To book a Boot Camp session, please contact overlapping with TV platforms?
Client Services Manager, Sarah Mowbray: Project Dovetail: How will it work and what will you get?

Reboot is designed to advance the skills of To book a Reboot session, please contact
those who are already established in the Insights Manager, Douglas Whelpdale: doug.
industry and use BARB data but are curious
to find out what more they can do. These

If you have any questions or would like to We hope you have enjoyed reading this
know more about other reports from BARB, report and we look forward to your feedback.
please contact
There is also plenty of information on our
website, which includes further details
about our plans for the future. In the About Broadcasters’ Audience
us section of our website you can also find Research Board
contact details for all members of the BARB
team. @BARBtelevision
To keep up to date with BARB and the latest
reports, follow us on Twitter and LinkedIn or
request to be added to our What’s new from
BARB mailing list.
W W W. B A R B . CO .U K
020 7024 8100