Insight

Welcome to the first 2022 edition of What People Watch, our occasional series in which we explore different aspects of how UK audiences are watching television now.

Understanding people is the central pillar of what Barb does. Our always-on reporting has continuously developed, allowing us to deliver greater insight on how viewing behaviours change as people have more options to choose between.

But sometimes, the more things change, the more they stay the same. Let’s take the viewing of Premier League football.

The recent upgrade in our daily audience reporting to include streaming services provides a great opportunity to see how audiences are watching Premier League games on a streaming service, as opposed to an established broadcast channel.

Amazon Prime Video first won rights to show Premier League football in 2018, airing their first match in December 2019. And now Barb can report how — or perhaps we should say if — the audiences Amazon attracts differ from those watching on Sky Sports and BT Sport.

Our first opportunity to analyse these audiences was presented almost as soon as we started reporting viewing to SVOD programmes. Amazon screened all ten Premier League games played between Tuesday November 30th and Thursday December 2nd 2021. The whistle-to-whistle, rather than whole programme, audiences can be seen in the table above.

As we often see from BT and Sky, games featuring Manchester United and Liverpool appear to attract the highest audiences when aired by Amazon.

In this instance we have a more direct comparison for the top-placed Amazon game in the table above with another Manchester United game also played on a weekday evening and aired by BT Sport. That game, on Wednesday January 19th 2022, saw Manchester United beat Brentford 3-1 and achieve a whistle-to-whistle audience of 997k (calculated in the same way as for the Amazon matches above).

It might be surprising that the audience to the match aired by what is primarily a shopping platform is almost double that on the established broadcast channel. Clearly the opposition is a factor; we have to accept that Arsenal are quite good at football, with a following to match.

The Manchester Uniter v Arsenal game would also be more attractive to neutral fans, featuring two teams who, allegedly, play some of the better football in the league. This game finishing 3-2 in Manchester United’s favour contained plenty of incident to keep audiences interested.

Moving 30 miles or so west, we take Liverpool as the next subject of our analysis as we compare the audience to the game versus Leicester on Tuesday December 28th 2021, screened by Amazon, and that for Liverpool’s 4-0 demolition of Arsenal which was broadcast by Sky Sports on Saturday November 20th 2021.

The Arsenal game achieved a live+VOSDAL programme audience of 1.72m across Sky Sports Main Event and Sky Sports Premier League on a TV set. Liverpool’s match playing away at Leicester City, by contrast, achieved an audience of 1.16m on Amazon.

In these two audiences we may we be seeing the impact of differing points in the year, week and time of day. The Leicester – Liverpool game was an evening kick-off and, just after Christmas, came at a time when we can anticipate that many extended families might still be gathered together. In such circumstances competition for the remote can be fierce and — outside of major tournaments — football is rarely the consensus choice.

At various different points during this match, non-football following residents in the home might have wanted to watch A Very British Scandal (BBC1), The Voice Kids (ITV), Extraordinary Escapes with Sandi Toksvig (Channel 4), or Traffic Cops (Channel 5), not to mention the likes of the ever-popular Jonathan Creek, which was being repeated on Drama during game time. The lower audience recorded for the game aired by Amazon has undoubtedly been impacted by such pressures.

The Arsenal game, on the other hand, was a relatively non-descript Saturday afternoon in the latter half of November. Those in the home who wanted to watch the football would likely not be competing with other family members for possession of the main TV set and certainly would not be competing against any live prime-time programming from other channels.

One benefit for the sports viewer of the rise in SVOD services’ popularity is that a live event can claim precedence over a programme that can be streamed at any time.

Taking these two Liverpool matches, it is also interesting to consider the extent to which the audience for the two screenings is shared. Clearly with different opposition for Liverpool in each game we would expect a portion of the audience to be unique, but audiences for matches featuring the better teams in the Premier League would also be expected to attract a large neutral following.

Comparing the two audiences we can see that 327k watched both games, accounting for 19% of the audience to the Liverpool – Arsenal game on Sky and 28% of the audience to the Leicester – Liverpool game on Amazon.

Finally, we can see if there was any difference in the audience profile for each match. There is almost no difference at a gender level. The audience is skewed male on both platforms, but very slightly more so for the Arsenal match on Sky at 70% male, compared to 69% male for the Leicester match on Amazon.

Similarly in terms of social grade we see little difference, although the audience for the match on Sky does appear to be slightly more upmarket – 64% ABC1 compared to 61% for the Amazon game.

The most marked difference we see is in the age breakdown of the two audiences. Those watching on Amazon are notably younger than those watching on Sky. 12% of the audience on Amazon are 16-24 compared to 6% of the audience on Sky, while 21% are 35-44 on Amazon compared to 18% for the game on Sky. Naturally, this means there are more older viewers on Sky – 22% are over 65, while only 16% of those watching on Amazon are in this group.

Perhaps more interesting is that the age profile on Amazon for this football match is quite different to the average age profile across the whole service for December 2021. Where we see 11% of the football watching audience were aged 25-34 this group made up 18% of Amazon’s audience across all content. And while the audience on Amazon was younger than that on Sky the proportion of those aged 65+ (16%) was far higher than the average of 9%.

Commentators, of the media rather than football variety, have often speculated that Amazon’s only real interest in providing video content is to sell more things. While providing a younger audience for football matches than Sky is clearly of interest to media types, the most interesting part for Amazon may well be pulling in older audiences who will then go shopping.

In this What People Watch we have focused mainly on games featuring a few of the Premier League’s most popular teams: Liverpool and Manchester United, as well as Arsenal. We have seen that the audience watching via Amazon is younger, very slightly less male and a little more C2DE. Although these balances do shift when we analyse other games.

There is also a significant shared audience between Sky and Amazon, as we might expect. Sports enthusiasts, regardless of the sport, are increasingly becoming accustomed to a need to seek out the platform that is hosting their game of choice. It makes sense that there is a core fanbase who will seek out their team on any platform. Our ability to analyse audiences across SVOD and broadcaster providers gives us a chance to understand these people.

We also see that, as with all programming, audiences will be impacted by the alternatives available on other channels at the same time. Not everyone appreciates the spectacle of 22 people chasing a small ball around a field. Some (in my household at least) even refer to all sport as ‘green-screen’.

The television viewing ecosystem may be changing dramatically, but arguments over the remote (assuming you can find it) remain. Some things never change!

A note on methodology:
As programme data for Amazon, Netflix and Disney+ are reported on an as-viewed basis, it is not straightforward to create a minute-by-minute viewing audience for a programme allowing for analysis of programme parts like our whistle-to-whistle figures in table 1.

We also cannot assume that viewers to a given service during a certain time period are watching the content we are interested in as we can with a broadcast channel.

To overcome this, we can create a dynamic target for viewers of the programme in question and then run minute-by-minute analysis at a service level, Amazon in this case, across the whole day. With a live event we know it is not possible that people were watching the action before a certain time. Therefore, we can be sure that the audience to the service as defined by our dynamic target were watching live, or close to live between certain times.

These figures can be considered comparable to live+VOSDAL figures as viewers who have started watching a game after kick-off on the day the game was aired will be included.

We should acknowledge, however, that not all VOSDAL viewing will be reflected in these minute-by-minute analyses. The Leicester – Liverpool game was a 20:00 kick-off. Our whistle-to-whistle audience figure stops at 21:56 – approximately 1 minute after the final whistle. Viewers who started watching the game after this point on day of airing would be included in live+VOSDAL figures in an as broadcast view of the data, but are excluded using this method.

This limitation is not unexpected given that we continue to monitor these services without their active participation. Were they to actively participate in Barb measurement it would be possible to provide data with greater accuracy.

Doug Whelpdale, Insights Manager, Barb