What might you use the Competition function for? You can probably think of a number of uses. I have gone for just one….

On January 29th, the BBC announced that Ripper Streetwould be renewed for a second season. I am going to try to show you how Competition analysis could be useful to the team that is starting to plan a second season.

It would  be good to know, as the team develop a new season of scripts, exactly which shows performed best and what can be learned from them for the new season.

There you have to be careful — because so many secondary factors. So one of things that this work can do is to prevent you from drawing incorrect conclusions.

(For those who do not know it, Ripper Street is crime series set in Victorian London and made by Tiger Aspect Productions, co-produced with Element Productions. It ran on BBC1 in the UK between Dec 30 and Feb 24.)

I am going  to that section of the AQUA Transmission Report called Competition.You can reach Transmission Reports by using Programme Search.

In the Competition section you can study the content that was transmitting on other channels at the same time as the content you’re most interested in was on air.

To start the process, I built this table from the data in AQUA so that we could compare  two different successive episodes: Jan 20 2013 and Jan 27, 2013. I have put a commentary in a panel alongside the figures — all this is information you can get from the report.

What conclusions can be drawn from this? The first thing you learn from this exercise is an old truth. Ripper Street is competing with another drama on ITV. 9:00pm is a prime time for drama on both ITV and BBC1 so this clash is hard to avoid. More often than not new mid-evening drama will split the audience. (Drama is, after all, the most-watched genre on TV.) But in this case Ripper Street outperforms Mr. Selfridge on ITV. So both episodes hold up well.

Does that mean the Rippers St episode on Jan 20 was “better” than the next one because it had 500 more viewers? No, not at all.  Wonders of Life was a stylish nature series from the BBC Natural History stable, fronted by Professor Brian Cox. It is potentially attractive to the same audience that watches BBC1. So the second episode may have lost some of its audience to BBC2 and Brian Cox. (Mr. Selfridge on the other hand seems unaffected.)

The schedule changes on Five are probably less relevant: while it lost audience after the end of Celebrity Big Brother, Five probably did not draw many viewers from BBC1.

We could have gone further looking at the Competition on smaller channels such as Sky1. It’s all available.

We have designed AQUA to be able to be used without too much training. The analysts at Attentional can provide more detailed analysis. For example they could tell you how many viewers of Ripper Street’s first episode did actually go over to Brian Cox’s Wonders of Life. Like all research, a first analysis will throw up questions that need further enquiry.

Careful research is a way of avoiding error. 

Attentional’s AQUA system has an easy option for this. You can go into Programme Search, put in the name and get all the editions of a series.

Then you can select the ones you want to run a Series Report on.

The table looks like this.

There is also a chart option.

These show that the series had a fairly steady performance, with a gradual decline across the series. The later episodes began to fall slightly below the channel benchmark.

But there are some things to watch out for when it comes to running a Series Report.

BARB data – it’s the same for other audience measurement systems using peoplemeters - comes through in waves. The first wave is generally called “overnights” in the UK – in the US sometimes “dailies”. They report minute by minute viewing figures. Agencies who publish these data, as we do, have to lay their own schedules against the data to get viewing figures for programmes. About a week later the programme timings are verified by broadcasters and shortly after that the commercial breaks are removed from the figures also.

There’s another thing. Lots of content is time shifted these days. That’s provided a challenge for TV audience measurement. The solution for the British provider BARB is as follows: the first dataset covers live viewing and content viewed on the same day – that dataset is called Live + VOSDAL – standing for Viewed On the Same Day As Live; the second dataset covers live viewing and any viewing of the same programme in the following week – that dataset is called Consolidated; the third dataset removes viewing data for commercial spots – that is called Programme Data, sometimes referred to as Gold Standard.

Notice that the table I have shown goes up to a date less than a week before this post was written on February 22nd. So the only data available is Live + VOSDAL.

The AQUA system offers the option to select the Data Type you want. It also offers a default option, labelled Automatic, which offers the best available set. If you select a Series Report it will only offer you data options that apply to all the programmes selected, since the report would be meaningless if different content was rated on different data.

Because I ran my first report on all episodes of Ripper Street, we weren’t able to measure time shift viewing for all episodes.

So I ran a second report excluding the most recent content to see how many people recorded Ripper Street and watched it in the same week: in other words, to see how much the content gained from time shifted viewing:

You can see from this how significant time shifted viewing now is.

This is a very simple approach. For a more sophisticated use of  series  information I strongly recommend Sarah Frost’s Screenwatch blog on Ant and Dec’s Saturday Night Takeaway published on February 25th. 

I showed how our Performance Index is a smart way to assess relative performance – something that can be hard to figure out if you only have an audience figure or a rating.

That’s because a programme should be assessed against how programmes of the same type usually perform on the same channel.

Where  would you go next? Probably to see what kind of audience the show got, that is, what sort of people were watching. The jargon word is: Demographics.

AQUA makes it easy to find out about the demographics of your audience. The tables below, which come from the same report on Ripper Street, segment the audience into demographic groups.

Here we use a different way of measuring relative performance. The “Benchmark” in the table below is more general: it tells us what sort of audience the channel would expect to get at this time. It is not specific to the genre of the content. 

The three columns in the table below measure the audience by age in different ways. The first, Profile, measures the proportion of the programme audience each demographic provides; the second, Share, measures the share of the people viewing at the time; while the third, TVR, measures the proportion of the population watching the show. (The abbreviation TVR, which stands for TV Rating. comes from advertising. Treat it simply as I have suggested – the proportion of all the people in the country in each group that watched your show.) 

Overall, Ripper Street got a somewhat higher share of people over 45 than usual for its BBC1 timeslot, and lower share of children and young adults. Nor were the over 65’s quite so keen, though the TVR measure tells that nearly one in 5 of all older adults watched the show. Most of the time the measures in the three columns confirm the same message. So, as a basic approach, just look along the columns to see the pattern. (We may look at rarer cases where this does not happen another time.) 

A second demographic table tells us that Ripper Street underperformed among the highest socio-economic group, which you might define as the professional classes. Although you cannot see this from the figures, it is quite easy to work out that the AB group is about 320,000 short of what might have been expected on BBC1 at this time. Tooltips in our reports explain the groups – they are based on the occupation of the main income earner in the household. If you are unfamiliar with these groups, you can learn more in the Glossary which is accessible from every page in a menu at the bottom of the page. Look for Demographics or Social Grade. 

On the whole – and unless a show is targeted at older groups – it is best not to have too much of a bias towards older groups. The main reason is quite simply that they watch more TV. So the decision to select a particular show is a less positive choice. On commercial channels which draw revenue from advertising, much content is targeted at households with young families, so, again, they are looking for shows that capture lots of younger adults. 

Young adults also tend to be busier mid-evening with family chores and children to get to bed. So a further question about this show is: did they record it and view it later? We will be able to look into that in later posts. 

You don’t use a service unless it helps you do a better job. And everyone in business knows you have to offer something your clients can’t get anywhere else. 

Sometimes the benefit is very simple, like getting accurate and important information very, very fast.

So that is where I will start, going into more advanced applications as the year goes on, introducing new developments as they come up. 

Let’s say you are part of the team that made Ripper Street  for BBC 1 in the UK. And let’s start with the first thing you probably want to know. How did it perform? (There is a channel executive on the line: you need a fast – but confident – reply). 

Ripper Street is made by Tiger Aspect , now part of Endemol, and by Lookout Point . Ripper Street is exec-produced by Greg Brenman, Simon Vaughan, Ed Guiney and Andrew Lowe.

The people at Tiger Aspect get our e-mail overnights every day, and some staff use our online research tool AQUA. Both are parts of our overnights.tv service package. I am going to use AQUA in these blogs because it has so much scope and flexibility.

Here is the summary they would have got on the morning of Jan 31. It tells us that Ripper Street got an audience of just over 6m viewers. The figures in our AQUA database take only a few second to retrieve:

You might want to evaluate that. Is this good? Is this just average? Outstanding or poor? You need a benchark.

You could look in our Top Programmes Report. But going to another report is a bit of a diversion.

No, you don’t need to do that. We have actually added a measure that gives you a fast readout on Ripper Street’s relative performance. It’s at the end of the summary data above and it’s called Performance Index. It’s a clever number because it compares your show with shows in the same genre in the same timeslot on the same channel over the last 12 months. A score of 100 means its average for the channel and genre. The first episode of Ripper Street wasn’t a blockbuster but it performed up to expectations.

Performance Index is unique to overnights.tv. The Glossary at Overnights.tv gives a longer description.

After that, there many more places to go and many other things you could find out about Ripper Street and its viewers. We will visit some of these in the next few blogs.

In the next one we might dig deeper into that audience. Who were they?

I was a producer once – and loved my job. I stopped being a producer over 20 years ago and still miss it. It’s hard to imagine now but the slot I produced on Nationwide back then, with Valerie Singleton and Richard Stilgoe, sometimes got 12m viewers.

I have always wanted to know just a bit more than anyone else about my craft. I have set up companies to get this knowledge.

Producers are still the people I know and understand best. I would like to see our services on every producer’s desk.

Are political opinions born or made? Are they determined as much by rational choice as by genes? Are political attitudes hard-wired?

For example, some twin studies show that identical twins, separated at birth, seem to end up with very similar political attitudes.

Bruce Goerlich, who is Chief Research Officer at Rentrak, has done an interesting thing in his blog. He separated Liberal and Conservative viewers in the US by selecting people who regularly watched Fox News or MSNBC. Rentrak can do this because that company collects data from 20 million set-top boxes around the US. (Hard to repeat, though, in other countries without avowedly “political” content.)

The results showed a country that was pretty polarised as you will see from the chart above. There were long tails in the distribution of content that mainly Conservatives – or mainly Liberals – watched. He gives examples.

Down the Conservative end we have NCIS: Los Angeles, Criminal Minds, College Football and NASCAR, and right down at the end of the long tail, Little House on the Prairie.

And Liberals? Modern Family, Family Guy, The Emmy Awards, The Office. And at the end of that tail Kate and Perle.

And what is left in the middle? You’ve got the Simpsons and Grey’s Anatomy, slightly liberal, and Survivor, slightly conservative.

Views doubtless change with age, and with education and other factors, too. Bruce could doubtless have cross-tabbed viewers by age and gender. That would give us another dimension.

Those of us who are in research business are all looking for ways to reduce uncertainty. So, of course, are political campaigners who may find it very useful to understand the viewing behaviour and cultural preferences of their voters.

I would like to propose another variable which could turn out to be useful.

Recently we have started to work on the issue of Temperament with help from psychologist Dr. Dario Nardi, who is expert on this subject. The Myers-Briggs system is widely used in profiling. There are 16 profiles in this system but on Dario’s advice we are working with a reduced subset called Temperament. There are four of these.

The proportions of the population for each Temperament in the US are well known. The actual proportions look like this: Stabilizers (45%), Improvisers (35%), Catalysts (12%) and Theorists (8%). So a large part of the US population is covered by two groups, Stabilisers and Improvisers. Yet apparently most of the creative people in TV are Catalysts.

Stabilisers feel natural conservatives with a small (c”). They like order, the respect traditional values, they like to know where they stand in a process. Improvisers like action, they take risks, they are impulsive.

All this means that a political party, like a TV channel, could investigate its audience on a number of dimensions, capturing preferences by age, gender, and temperament. That could offer a range of categories or subgroups on a number of dimensions,  – Liberal 18-49 Stabilizers and so on.

And because “values” are hard to capture in qualitative surveys, the content people choose could well turn out to be informative, a good guide for both schedulers, looking to hold audiences through an evening, and political advisers, looking for an even better understanding of their supporters.

There’s actually some good history around at present. I liked Nutopia’s The British a lot. But Untold History, narrated by Oliver Stone, is something different again. It’s a reminder that every nation’s history, the one that’s taught in schools, is in the main just a pretty story. British kids do not get told about the Opium Wars in the 19th century when we forcibly prevented the Chinese from dealing with the scourge of drug addiction that was ruining their country.

If you had problems with the video here is a link to the trailer.

On a completely different theme there’s a new drama series on MTV India called Bring on the Night.

It’s about four guys trying to set up a club together, and it’s heart-warming and funny, very different, with some darker moments and a wonderfully humorous look at Indian public officials. There is one British guy in the group of friends called Patrick. Here is the trailer and here’s the latest episode on MTV’s site — it may be taken down after a few days though.

Both shows have made me think about globalisation.

Anyone who works in news or current affairs knows, on the whole, where most of the audience is – looking around its own backyard. No point in trying to duck that. Human beings are tribal.

But globalisation is the key to the future of the planet. A joint recognition that we will have to manage the future of this planet together is crucial.

Yet the processes of globalisation are tentative, uncertain, deeply bureaucratic, perhaps necessarily so. We cooperate in our own tribe — more or less — and look suspiciously at other nations, — that’s if we are not at war with them as we probably have been for most of our evolutionary history. Few people have the stomach for the impersonal mechanisms of globalisation. I admire those who have stood out for it.

I admire the way Ted Turner became a champion of the UN after starting CNN, and donated a billion dollars to it.

Globalisation, which is after all just about seeing the big picture, has its heroes — like George Marshall, who helped put the Western European countries back on their feet after WW2. (He’s definitely a hero of the Untold History.)

Countries have open and closed phases. When we were in China, we found people curious and interested and very hospitable.

I admire the Nobel Peace Prize Committee for awarding the prize to the EU. No, the EU probably did not guarantee peace in Europe. It took American and British troops stationed in Europe for 30 years to help do that. But the EU has brought over 30 different language and cultures, i.e. more than 30 different tribes, together to create a union of nations that is unprecedented in the history of the planet.

People can help make globalisation real. Matt Graham has learned a history different from the one he learned in school and mastered it well enough to write a series about it. Patrick is acting in a comedy series about four friends that’s half in English and half in Hindi.

Both of them happen to be our sons and I am very proud of what they’re doing.

Globalisation has to recognise all the instincts of loyalty and patriotism that most of us feel for our own countries. But crossing over, getting out, entering other cultures is so important. People who do that help to make globalisation work – along with the hard, tedious politics of course.

But for those who believe that globalisation has to be embedded in our cultures to work, we still have a long way to go. Most of our entertainment is pretty provincial. let’s face it.

The Entertainment Master Class is part of the Entertainment Academy, “a think tank created for the industry by the industry”.

One of the things we went through is what I called: “Television’s four most important numbers”. They capture the ways in which a one-time homogenous stream of content, viewed only in one place at one time, is splitting into different streams – some of which have, as it were, gone underground and become invisible.

Head of Programme, John Gough, introducing a session.

Television viewing has traditionally been accurately measured. We knew where viewers were, what they were doing, and how many of them were watching. Uncertainty is new to it.

Now we are not so sure about any of these things. Estimates replace hard numbers.

Here are four categories:

• Live Viewing: can be monitored accurately using standard peoplemeter systems like Nielsen or BARB. It can also be measured very accurately at household level at the headends of cable or satellite systems.
• Time shifted Viewing: monitored accurately if viewed shortly after recording on a device like Tivo or Sky+.
• Long-form TV content on the Internet, streamed or downloaded: hard to measure but enough data for reasonable estimates. At Attentional we use ad hoc systems building on various datasets.
• Long-form content on other media (e.g. DVD’s): very hard to monitor. Archives of coded content are being built that will enable most content to be recognised one day.

I asked the group I was working with to think about their own habits and come up with estimates of their own viewing, using average minutes of viewing per day.

They came up with average numbers like these:

• Living Viewing: 100 minutes per day
• Time-shifted Viewing: 60 minutes per day
• Internet Viewing: 30 minutes per day
• Other media (e.g. boxed sets): 30 minutes per day. (Big variations here: there were parents with small children in the group!)

I then told the group what, to the best of our knowledge, the average times were for a population as a whole. I had to use UK figures here — there are quite big variations between countries, with the UK on the high end.

The figures are:

• Live Viewing: 220 minutes
• Time shifted Viewing: 20 minutes
• Internet Viewing: 20 minutes.
• Other media: not known.

Note the variation: much less live viewing and much more other viewing from the Entertainment Masterclass when compared to the population as a whole.

What accounts for it? This was an audience of young tech-savvy TV producers and executives who are interested in, and familiar with, smart digital devices.

But we know that is not the whole story because other similar surveys have been done with members of the general public. When you do such surveys, you get a similar result. People overestimate both the time-shifted part and the internet part relative to actual behaviour captured on reasonably reliable devices.

Why? Perhaps because it is cooler and feels more sophisticated if you watch and use a recorder and skip the ads. Perhaps because people record more content than they ever watch and conflate recording with watching.

Or is it because that is ideally how people would like to view – in the face of technology that in most homes is not quite ready yet? In other words, the desire is there – but the devices are not smart enough.

This is an important psychological point. The desire could be a driver.

In which case there could be quantum jump some time up ahead when some smart technology innovation – like the IPhone – or a technology leap – like a rapid increase in broadband speeds – kicks off a new phase.

In the UK live viewing is dropping, each year, by two or three minutes a day. That is a slow rate of change. Young people watch less live TV, and show a greater appetite for digital devices, but the rate of change is not significantly higher.

Speaking ahead of Mipcom, Simon Cowell is reported to have said of YouTube: “it won’t be long before the videosharing site is a serious alternative to traditional channels…. When we reach that point, they’re going to be serious competition.”

We have long argued that YouTube has that potential.

But we also agree with Cowell that people want to watch their favourite shows on a plasma screen not a 16” monitor.

I remain surprised that Google has not used its massive financial resources to secure more attractive exclusive content. I am sure there are some very interesting conversations going on in Google boardrooms. Perhaps Simon Cowell is hinting that he wouldn’t mind a conversation!

The winning players will be the ones that get the timing and the technology right.

It got a lot of viewers. In fact they peaked for that race  and fell away afterwards.

Source: Attentional, BARB

Extended live Olympic coverage was something different for BBC1. So I wondered how its audience had changed. I did that by creating ratios, for each age group, between BBC1’s usual share and its share that night. Here are the uplift ratios:

Source: Attentional, BARB

So the uplift  came from young men and women. It was strongest for those between 25 and 34. Among people over 55 there was hardly any uplift at all.

The ratio for Men was 2:1 but for Women there was no uplift. It was mainly men,  then,  who came over from other channels to watch BBC1.

Source: Attentional, BARB

You are not surprised. I hear you saying : “Young men like Sport”, “Michael Phelps was already a Celebrity”.

From a scientific point of view these are descriptive statements. They explain nothing.

On August the 3rd, Rebecca Adlington competed in the 800 metres freestyle and was expected to do well, perhaps get gold. The audience peaked for her race too.

Source: Attentional, BARB

This time the audience uplift for BBC1 did not come predominantly from young men and women between 25 and 34. This time it was across the board. Even older adults were there in numbers.

Source: Attentional, BBC1

And the ratios were different too: more women this time.

Source: Attentional, BARB

You might find this unsurprising and, again, not particularly useful.

But let’s step behind the obvious and ask what is going on in the psyches of the millions of people watching.

We watch winners because we all want to win — in our own lives — and we want our tribe to win.

Athletes are  fit people. That same word, “fit”, is, as it happens, also used in evolutionary science to mean individuals who are adaptively fit, that is, likely to pass good genes onto the next generation. So “Fit” means successful, powerful, attractive people — we call them “fitness models”.

Humans instinctively pay attention to fit people. We watch winners because we all want to win. We  look to see  how the winners and losers handle themselves before and after the race as well.

On balance, men are slightly more interested in physical competition than women.

Men watched the Phelps race to see  a fitness model of their own generation and because they like competition. Women watched in smaller numbers —  in the Mens’ race, there was no same-sex fitness model for them. But that changed for the Adlington race. It got to women too.

What about the older viewers? Familiarity is more important to older than younger people. Older British viewers would not have  known Phelps that well and would not feel close to him as an American. Adlington on the other hand was familiar to them, part of their “tribe”. So Adlington got a wider span of viewers.

We invariably back our own tribe. Competition between nations is intense. We respect and admire great competitors from other tribes, but there is not the same urgency  about wanting them to win.

At the Olympics each nation gets new heroes to admire and follow. Over the years TV has taken us closer and closer to the winners and losers as they leave the ring, track or pool, jubilant or shattered.

For the British the new heroes are Mo Farah, Jessica Ennis, Laura Trott, and  many more. For America and others nations the  line-up of new heroes is a different one.  Katie Ledecky, the teenager who beat Adlington, is famous now, and other teenagers will be watching her every move.

Some of you who write or create stories will say: “I want to know more”.

If so, ask us about workshops. Last week I was in Israel for the Entertainment Master Class, which brings together producers from all over the world.

Or if you want a good read try the recent book by Jonathan Gottschall called The Storytelling Animal.

Our knowledge about stories and the human brain is moving forward at a rapid rate.

One recent discovery, the Mirror Neuron, has huge implications for our understanding of how the brain deals with stories.

The Mirror Neuron was discovered by an Italian team in the 1990’s who found that when macaque monkeys observed an action, the same neurons fired in both the actor and the observer.

We now know that happens with humans too. Scientists are asking if this explains the phenomenon of empathy. A drama cannot succeed without empathy for one or another protagonist. It’s empathy that makes us care whether the heroine gets the man she loves or saves the child.

Here is one scientist describing what happens in his view when mirror neurons fire:

There is a dictum in Neuroscience: “if it fires, it wires”. So are stories actually wiring our brains?

Those who commission, pay for or publish stories are taking a huge bet on audience preferences.

Anything that reduces that risk is going to be highly valued.

Six years ago we were in very new territory. We studied scientific knowledge available at the time, knowledge about how the brain processes stories, work on how and why humans have certain preferences.

We were cautious, but we built an evaluation framework based on that knowledge and have continued to refresh it.

It has worked well and proved effective in predicting the success of the scripts and pilots we reviewed. It has also proved successful in identifying the key elements of successful entertainment shows.

Dario Nardi speaking

We are moving on now – into a new area. We are pleased to announce that Dario Nardi is soon to be joining our team as Scientific Advisor.

Dario will add an exciting new dimension to our work. He is an expert in Temperament Analysis.

We know peoples’ appetite for stories varies along a number of dimensions, like age and gender. We think Temperament may be another significant variable in predicting how a given property will perform. Slightly different brains probably respond in different ways.

That could provide yet another criterion for choosing the content that justifies investment or development.

Zhang, Ping (Apple), Zak Shaikh and Olivia Song Prepare for a Session

Beijing airport is a complex of what look like huge scallop shells, betokening a vast airport.

We have two days there before moving on to our client’s location, Hunan Province, whose capital, Changsha, is three hours by air from Beijing.

We are giving three days of workshops, explaining our approach to the evaluation of content and illustrating it from a wide range of case studies of new entertainment and drama from the West.

We also give a session on business models from the West and examples of new research techniques.

Our party consists of Zak Shaikh, Head of Content, who has travelled from the US with Olivia Song acting as translator. Our client is Yike Tang, Deputy Head of Entertainment, at Show TV.

Zak has assembled his case studies into genre clusters with many subgenres. He covers and enormous number of cases. Why, for example, has Come Dine with Me been so successful? That’s one example where cultural differences do come in — “dinner parties” are a Western thing.

I present the shorter Science of Entertainment modules. These present the latest knowledge we have been able to gather on the application of neuroscience to entertainment. We believe this is a very important trend, and a very important part of our services.

We also present our past and future Research and Development plans.

Our audience is particularly interested in a new approach using Temperament analysis which we will introduce shortly. We hope to begin developing models for applying this technique this summer, working with UCLA professor, Dario Nardi.

ShowTV has the specific task of innovation: many of its most successful shows go on to Hunan Satellite TV, which is networked across China and which became famous for the big hit show known as Supergirl.

Supergirl was China’s first really successful talent show.

It proved so successful that Hunan Broadcasting became the leading provider of entertainment. Its satellite channel won huge audiences nationwide.

But Supergirl was cancelled in 2011 on the orders of the regulator. Hunan Entertainment now has to limit the number of talent shows in its primetime schedule.

However, ShowTV, our client, is a regional service, and not bound by the rules of its larger sister. It can continue to play reality entertainment in the primetime slot. That means it can carry on innovating and creating new formats. ShowTV is a real hothouse, with a very young team, keen to do new things.

Our Team with Yike Tang and the ShowTV producers

The sessions are very lively. In the evening over dinner we discuss their programmes with the producers and try to suggest improvements.