About About Contact Contact

Who's watching TV...

...where, when and what? By any measure television is still the dominant medium of the digital age in the UK and U.S.  It is the primary source that people use when they wish to be entertained.  Let me justify this: according to Nielsen, the company that governs the measurement of TV watching behaviour in the U.S., the average American spends 35 hours per week watching video across all their screens compared to an additional five hours per week using the Internet on a computer.  And the U.S. Census Bureaux agrees:  According to the recorded census forms, the most popular leisure activity in America is watching television.  As people dedicate so much of their precious time to the television, it seems curious that many of them seem to be totally unaware of what determines which TV programs are made.   Here, sadly, we’re not talking about the creative angst of broadcasters, or the intelligent enthusiasm behind original ideas, or even the imaginatively juggled budgets underlying excellent programs.   No.  The bulk of the TV programmes that are chosen to be produced, and subsequently watched, are entirely shaped by what advertisers will buy - and that is determined by how television viewing behaviour is measured.   It’s all down to a continuous metric feedback loop - so whilst the mediocre won’t conquer the earth, they will conquer the satellites. Audience measurement is so central to the business of TV advertising that advertisers won’t pay money without the illusion of buying a quantified audience, provided by the metrics.  These apparently explain the numbers of people who will see, or will not see, the advertising interspersing the programmes.  So, according to audience metrics, television ads are unavoidable.  In the UK, even paying a subscription to the satellite provider Sky will not prevent the viewer being exposed to TV advertisements.  But in the UK there is one exception to this - the government collects compulsory revenue for a television licence and uses this money to fund the BBC, which has a remit to transmit programmes without commercial advertising.  Nevertheless, much of what the BBC airs is still shaped by the needs of advertisers.  This happens in two ways, firstly, the BBC buys in series made in the U.S. and these, of course, have been deliberately designed to attract specific target audiences for advertisers and therefore act as advertising vehicles.  Secondly, many BBC programmes are created as a competitive response to those shown on commercial channels which are funded by advertising. At the moment there are some fascinating changes happening within the TV industry and these changes, at a fundamental level, all revolve around how the sound and images that make up a television image get delivered to a screen.  As TV viewers respond to the possibilities presented by having access to programmes on multiple devices, and being able to choose what, when and how they watch, it becomes obvious that the television industry is dealing with a fractured audience measurement system.  The Internet, or more precisely the Web, is now acting as the delivery mechanism for a growing proportion of the output of the TV production industry, and this change is radically altering the measurement status quo upon which the whole television business is based. It used to be that all TV sound and imagery was transmitted across “the airwaves” from tall masts just like the very first transatlantic TV broadcast by John Logie Baird in 1928 made shortly before  the very first BBC television programme was transmitted.  TV programmes still arrive using “the airwaves” for many people around the world, but now, more often than not, it is a digital signal that gets received, of which a good proportion will have been sent from a satellite.  For the original television measurement business this situation was relatively straight forward:  Create a sample of people, (in reality a surprisingly small sample because of the expense), and equip their TV sets with a device that records when the television is on and what channel is active.  In 2009 in the U.S. that sample was 0.02183% of the total population.  Get the individuals in the families that make-up the sample to record in a diary when they watch television and the combined result is Nielsen Ratings The “ratings” of a TV show are vital as they determine whether a show lives or dies, and how much the “channel” showing it can get paid for the advertising displayed as the programme gets “aired.”  Before the phenomenon of the fragmented TV, the only other change in the method of delivering television programmes to a domestic environment happened in the 1970s with the development of cable TV.  But the method of measurement didn't really change; the monitoring device was merely expanded to cater for the extra channels that become available via the cable.  Later, the introduction of digital equipment, such as Digital Video Recorders (DVRs), enabling viewers to time- shift their viewing, started to upset the television metric apple cart.  Nielsen's response, along with other companies in the same business, was to elicit more diary entries from a sub-set of homes equipped with their tracking devices who owned DVRs.  At that stage of the proceedings it was becoming apparent to most people in the business that the all-important TV measurement system for advertisers, television producers and TV stations was creaking and really showing its age. Then suddenly, in 1994, as the public use of the Internet was growing exponentially, a new way of delivering moving images and sound to a domestic television screen was introduced.  Not via “airwaves,” or a thick, dedicated cable, but by using the already established infrastructure of the telephone system using much thinner wiring.  At first thought the Internet seems an unlikely method of delivering moving images and synchronised sound effectively because they arrive via a distributed network in packets of discrete data that have travelled different routes.  These packets have to be reassembled into the correct order before they are processed onto the screen and sound system.  In the beginning this was indeed a problem with slow dial-up Internet speeds, but when broadband and faster network technology came along this difficulty was essentially solved.  The result is that we now have a growing number of Internet-connected TVs, something Samsung brands as “Smart TV.”  Television viewing is also changing fundamentally because faster Wi-Fi and other technologies mean that programmes can be viewed on mobile devices anywhere.  Part of the reason that the changes in TV watching behaviour appear simpler than they actually are is down, yet again, to the way that television behaviour still gets measured.  Small sampling systems like Nielsen panel data can no longer cut it in a “big data” world. The numbers used in the chart above come from the BBC's data-warehouse and are as comprehensive a measurement of actual Internet viewing behaviour as you are likely to get at the moment.  And as the BBC is a publicly-funded body, and therefore agnostic about devices and platforms, this is unbiased data unlike much that is used in the television trade.  I've re-presented and edited the data to highlight some of the salient features.  First, some qualifying points about the data: these are only requests for full length programmes and that includes on-demand (streamed live) and catch-up (played within seven days).  The data, obviously, is only for the UK but it represents a sample size of 25% of the entire UK online population, exponentially larger than Nielsen’s sample sizes.  The BBC is working towards providing a 100% sample size but that is some way off at the moment.  In the chart on the left I have used the number of programme requests for the month of August for the last four years.  Keep in mind that the August 2012 figures will be somewhat inflated because of the Olympics.  The reason I’m using the month of August is that it is the most up-to-date data available.  Even discounting the Olympic surge, there is a clear and strong upward growth trend for radio as well as television audiences.  (For the statistically minded the growth pattern for both is very similar, with a correlation coefficient r of 0.86).  People now appear to prefer listening and watching their radio and television programmes when they want to, and not necessarily when those programmes are slotted into a broadcaster's schedule.  We are witnessing the slow death throes of “appointment” radio and TV as online access accounts for nearly 200 million programme requests within the month.  Each programme should be considered as long- form audio or video as the majority of the programmes will be in excess of 25 minutes duration and many will be around 50 minutes in length.  The unequivocal conclusion is that online access is capturing a substantial amount of available audience time. The really interesting aspect of this dataset is the granularity obtainable which shows the proportion of requests by devices used to view television programmes.  As changes in watching behaviour are happening so quickly I've sliced the August data to compare last year with the current year.  Isolating the data in this way makes it easier to see which devices are growing in popularity and which are declining.  The basic problem with Nielsen's measurement, or any other company’s measurement system, is that they are based on sample panels.  These panels are too small to provide any idea of the multiplicity of ways people are using technology to view their desired TV content.  By obdurately continuing to mainly rely on panel data to measure audiences, the advertising and television industries are continuing to shape programmes that are totally out of alignment with the behaviour of the viewers they seek to reach.  Look at the data:  Requests from TV platform operators (satellite and cable operators) are significantly declining and, unfortunately for Microsoft, requests from games consoles are also beginning to decline.  Using computers to view television programmes is also beginning to decline while tablet and mobile device requests are growing.  We are entering a time when more TV programmes are being viewed by individuals on their personal devices than ever before, and less viewing is taking place in a family group or with friends.  MyTV, in every sense of the word, is increasing.  Perhaps the biggest surprise is the relatively small number of requests from Internet- connected televisions.  According to much panel and questionnaire-based market research, there are around four million connected TVs in the UK.  Research shows that watching catch-up television is the most popular reason for buying a connected TV even though only 40% in the UK are connected to the Internet.   As the chart demonstrates, either (heaven forfend!) numbers are being overstated by market research companies, or some very strange viewing behaviour is going on.  If there really are around two million connected “connected” televisions in the UK, it seems odd that their owners only account for 2% of the requests for BBC programmes.  As requests from platform providers are declining it’s unlikely that “connected” connected TV viewers are using this route to view catch-up programmes.  One possible explanation for the disparity is that the “smart TV” interface on connected televisions is encouraging YouTube viewing and not the use of the BBC iPlayer.  Now that’s something for Nielsen to explore. November 2012
Click here to download the PowerPoint chart: Click here to download the PowerPoint chart:
...with analysis & insight...
Home Home Archive: Free PowerPoint download Free PowerPoint download
Click image to enlarge
Click here to download the PowerPoint chart: Click here to download the PowerPoint chart: Click here to download the PowerPoint chart: Click here to download the PowerPoint chart: Click to return to page Click here to download the PowerPoint chart: Click here to download the PowerPoint chart:
Archive Archive
About
Contact
Home
Click to return to page

Who's watching TV...

...where, when and what? By any measure television is still the dominant medium of the digital age in the UK and U.S.  It is the primary source that people use when they wish to be entertained.  Let me justify this: according to Nielsen, the company that governs the measurement of TV watching behaviour in the U.S., the average American spends 35 hours per week watching video across all their screens compared to an additional five hours per week using the Internet on a computer.  And the U.S. Census Bureaux agrees:  According to the recorded census forms, the most popular leisure activity in America is watching television.  As people dedicate so much of their precious time to the television, it seems curious that many of them seem to be totally unaware of what determines which TV programs are made.   Here, sadly, we’re not talking about the creative angst of broadcasters, or the intelligent enthusiasm behind original ideas, or even the imaginatively juggled budgets underlying excellent programs.   No.  The bulk of the TV programmes that are chosen to be produced, and subsequently watched, are entirely shaped by what advertisers will buy - and that is determined by how television viewing behaviour is measured.   It’s all down to a continuous metric feedback loop - so whilst the mediocre won’t conquer the earth, they will conquer the satellites. Audience measurement is so central to the business of TV advertising that advertisers won’t pay money without the illusion of buying a quantified audience, provided by the metrics.  These apparently explain the numbers of people who will see, or will not see, the advertising interspersing the programmes.  So, according to audience metrics, television ads are unavoidable.  In the UK, even paying a subscription to the satellite provider Sky will not prevent the viewer being exposed to TV advertisements.  But in the UK there is one exception to this - the government collects compulsory revenue for a television licence and uses this money to fund the BBC, which has a remit to transmit programmes without commercial advertising.  Nevertheless, much of what the BBC airs is still shaped by the needs of advertisers.  This happens in two ways, firstly, the BBC buys in series made in the U.S. and these, of course, have been deliberately designed to attract specific target audiences for advertisers and therefore act as advertising vehicles.  Secondly, many BBC programmes are created as a competitive response to those shown on commercial channels which are funded by advertising. At the moment there are some fascinating changes happening within the TV industry and these changes, at a fundamental level, all revolve around how the sound and images that make up a television image get delivered to a screen.  As TV viewers respond to the possibilities presented by having access to programmes on multiple devices, and being able to choose what, when and how they watch, it becomes obvious that the television industry is dealing with a fractured audience measurement system.  The Internet, or more precisely the Web, is now acting as the delivery mechanism for a growing proportion of the output of the TV production industry, and this change is radically altering the measurement status quo upon which the whole television business is based. It used to be that all TV sound and imagery was transmitted across “the airwaves” from tall masts just like the very first transatlantic TV broadcast by John Logie Baird in 1928 made shortly before  the very first BBC television programme was transmitted.  TV programmes still arrive using “the airwaves” for many people around the world, but now, more often than not, it is a digital signal that gets received, of which a good proportion will have been sent from a satellite.  For the original television measurement business this situation was relatively straight forward:  Create a sample of people, (in reality a surprisingly small sample because of the expense), and equip their TV sets with a device that records when the television is on and what channel is active.  In 2009 in the U.S. that sample was 0.02183% of the total population.  Get the individuals in the families that make-up the sample to record in a diary when they watch television and the combined result is Nielsen Ratings.  The “ratings” of a TV show are vital as they determine whether a show lives or dies, and how much the “channel” showing it can get paid for the advertising displayed as the programme gets “aired.”  Before the phenomenon of the fragmented TV, the only other change in the method of delivering television programmes to a domestic environment happened in the 1970s with the development of cable TV.  But the method of measurement didn't really change; the monitoring device was merely expanded to cater for the extra channels that become available via the cable.  Later, the introduction of digital equipment, such as Digital Video Recorders (DVRs), enabling viewers to time-shift their viewing, started to upset the television metric apple cart.  Nielsen's response, along with other companies in the same business, was to elicit more diary entries from a sub-set of homes equipped with their tracking devices who owned DVRs.  At that stage of the proceedings it was becoming apparent to most people in the business that the all-important TV measurement system for advertisers, television producers and TV stations was creaking and really showing its age. Then suddenly, in 1994, as the public use of the Internet was growing exponentially, a new way of delivering moving images and sound to a domestic television screen was introduced.  Not via “airwaves,” or a thick, dedicated cable, but by using the already established infrastructure of the telephone system using much thinner wiring.  At first thought the Internet seems an unlikely method of delivering moving images and synchronised sound effectively because they arrive via a distributed network in packets of discrete data that have travelled different routes.  These packets have to be reassembled into the correct order before they are processed onto the screen and sound system.  In the beginning this was indeed a problem with slow dial-up Internet speeds, but when broadband and faster network technology came along this difficulty was essentially solved.  The result is that we now have a growing number of Internet-connected TVs, something Samsung brands as “Smart TV.”  Television viewing is also changing fundamentally because faster Wi-Fi and other technologies mean that programmes can be viewed on mobile devices anywhere.  Part of the reason that the changes in TV watching behaviour appear simpler than they actually are is down, yet again, to the way that television behaviour still gets measured.  Small sampling systems like Nielsen panel data can no longer cut it in a “big data” world. The numbers used in the chart above come from the BBC's data- warehouse and are as comprehensive a measurement of actual Internet viewing behaviour as you are likely to get at the moment.  And as the BBC is a publicly-funded body, and therefore agnostic about devices and platforms, this is unbiased data unlike much that is used in the television trade.  I've re-presented and edited the data to highlight some of the salient features.  First, some qualifying points about the data: these are only requests for full length programmes and that includes on-demand (streamed live) and catch-up (played within seven days).  The data, obviously, is only for the UK but it represents a sample size of 25% of the entire UK online population, exponentially larger than Nielsen’s sample sizes.  The BBC is working towards providing a 100% sample size but that is some way off at the moment.  In the chart on the left I have used the number of programme requests for the month of August for the last four years.  Keep in mind that the August 2012 figures will be somewhat inflated because of the Olympics.  The reason I’m using the month of August is that it is the most up-to- date data available.  Even discounting the Olympic surge, there is a clear and strong upward growth trend for radio as well as television audiences.  (For the statistically minded the growth pattern for both is very similar, with a correlation coefficient r of 0.86).  People now appear to prefer listening and watching their radio and television programmes when they want to, and not necessarily when those programmes are slotted into a broadcaster's schedule.  We are witnessing the slow death throes of “appointment” radio and TV as online access accounts for nearly 200 million programme requests within the month.  Each programme should be considered as long- form audio or video as the majority of the programmes will be in excess of 25 minutes duration and many will be around 50 minutes in length.  The unequivocal conclusion is that online access is capturing a substantial amount of available audience time. The really interesting aspect of this dataset is the granularity obtainable which shows the proportion of requests by devices used to view television programmes.  As changes in watching behaviour are happening so quickly I've sliced the August data to compare last year with the current year.  Isolating the data in this way makes it easier to see which devices are growing in popularity and which are declining.  The basic problem with Nielsen's measurement, or any other company’s measurement system, is that they are based on sample panels.  These panels are too small to provide any idea of the multiplicity of ways people are using technology to view their desired TV content.  By obdurately continuing to mainly rely on panel data to measure audiences, the advertising and television industries are continuing to shape programmes that are totally out of alignment with the behaviour of the viewers they seek to reach.  Look at the data:  Requests from TV platform operators (satellite and cable operators) are significantly declining and, unfortunately for Microsoft, requests from games consoles are also beginning to decline.  Using computers to view television programmes is also beginning to decline while tablet and mobile device requests are growing.  We are entering a time when more TV programmes are being viewed by individuals on their personal devices than ever before, and less viewing is taking place in a family group or with friends.  MyTV, in every sense of the word, is increasing.  Perhaps the biggest surprise is the relatively small number of requests from Internet-connected televisions.  According to much panel and questionnaire-based market research, there are around four million connected TVs in the UK.  Research shows that watching catch-up television is the most popular reason for buying a connected TV even though only 40% in the UK are connected to the Internet.   As the chart demonstrates, either (heaven forfend!) numbers are being overstated by market research companies, or some very strange viewing behaviour is going on.  If there really are around two million connected “connected” televisions in the UK, it seems odd that their owners only account for 2% of the requests for BBC programmes.  As requests from platform providers are declining it’s unlikely that “connected” connected TV viewers are using this route to view catch-up programmes.  One possible explanation for the disparity is that the “smart TV” interface on connected televisions is encouraging YouTube viewing and not the use of the BBC iPlayer.  Now that’s something for Nielsen to explore. November 2012
Click here to download the PowerPoint chart: Click here to download the PowerPoint chart: