It’s finally happened, in late September 2018 BARB released the first instalment of Project Dovetail, with census level device-based data (which BARB has been publishing separately on their website since September 2015) being fused with people-based BARB panel data to provide demographic level audience estimates for TV viewing on PCs, Tablets and Smartphones, though viewing data for Smartphones is currently only available at the Individuals 4+ level. The main takeaways are that while multiscreen viewing (as was already clear from the historic device-based data) at best currently only adds around 1.5% to schedule-based TV viewing, its impact is more pronounced for younger demographics, and for the occasional programme it can be very significant indeed. Nevertheless, it seems clear that when it comes to high quality long-form video content, the key battle for viewers over the next 5 to 10 years will be fought on our ever bigger and smarter TV screens.
This is emphatically not something that has been lost on the FAANGs, who, unencumbered by the kind of regulatory scrutiny that a decade ago killed Project Kangaroo (a pay-VOD joint venture between BBC, ITV and Channel 4), have worked hard to get their apps on the home screens of as many Smart TVs and TV connected devices as possible. Netflix has even managed to go one further by striking deals with several major television manufacturers to have handy Netflix buttons installed on their remotes, and this is on top of the deals with pay-TV operators that make Netflix readily available through Virgin Media, BT(YouView) and, most recently, Sky Q set-top boxes. It should come as no surprise that globally 70% of Netflix viewing is through connected TVs. The net result, as outlined in my Blog from earlier this year, is that while our total TV screen time has hardly changed in recent years, there has been significant growth in Unknown viewing (which includes TV-based viewing of SVOD services) at the expense of schedule-based TV viewing, particularly for the younger age-groups.
In view of such stiff and largely unregulated competition, and with Ofcom’s EPG prominence review now well underway, it is hardly surprising that the PSB broadcasters have been lobbying hard for the Government to introduce legislation that ensures PSB prominence rules remain relevant. The case is being made that the value of their long established right (in return for their public service commitments) to the first 5 slots on all linear EPGs, though still extremely important (as I have argued repeatedly), is being steadily eroded, and so they are pushing for PSB prominence requirements to be extended to cover things like home screens and recommendation algorithms. The non-PSB commercial broadcasters naturally disagree, with Sky making the case for not extending the PSB prominence regulations, though it is perhaps telling of the likely trajectory of future legislation that in a recent speech Sharon White, Ofcom’s CEO, as well as reiterating her call for more collaboration, also noted that: “The ability to stumble across great UK shows, that an algorithm might not suggest, seems a principle worth protecting in an online world. […] Who knows. A ‘PSB button’ might be vying for space next to Netflix and YouTube on the television remote control.”
So, how much of a future threat are the SVOD services to schedule-based television? One thing that has caught many commentators by surprise is the sheer speed with which the SVOD services have managed to proliferate through the UK market. As part of its establishment survey, BARB has been estimating quarterly SVOD take-up since the start of 2014, and the occasional mid-year slowdown notwithstanding, the proportion of UK households subscribing to at least one SVOD service has increased rapidly, nearly tripling from 14% in 2014-Q1 to 41% in 2018-Q3. In less than 5 years SVOD has gone from niche to mainstream, an impressive achievement by any standards, particularly when one considers the fact that around 14% of UK households (a third of SVOD homes) currently subscribe to 2 or more SVOD services. In fact, as highlighted in Ofcom’ s Media Nations 2018 (pp. 12-13) report, purely in terms of the number of subscriptions, SVOD now exceeds pay-TV in the UK, as very few households subscribe to more than one pay-TV service. It is also highly likely that SVOD take-up in the UK will continue to grow in the coming years, with the leading incumbents (Netflix, Amazon and Now TV) continuing to invest in original content. The upcoming launch of major new SVOD services from Disney (set for a late 2019 US launch, followed by an international rollout) and potentially Apple (currently still only a rumour, though a 2019 launch seems probable), will in all likelihood also convert additional households to SVOD, as well as giving the incumbent players a run for their money. Encouraged by Ofcom, the UK’s PSB broadcasters have also been actively considering their own major SVOD propositions for the UK market, though whether this would ultimately be in the form of a joint-venture (potentially along the lines of the BBC/ITV BritBox service currently available in the US and Canada) or as independent services isn’t yet clear.
So, to what extent will we be even more of an SVOD nation by the mid-2020s, given that 41% of UK households currently have at least one SVOD subscription, how much longer can the rapid rise of SVOD really continue before it levels off? Ideally, we would answer this question by building a complex econometric model, accounting for potential causal factors like: the growth in the number of available SVOD services, how much they cost, how much they spend on content, the type of content they offer (is a significant move into sports rights on the horizon?), household demographics (which households are likely to be SVOD resistors), and not least the quality and availability of free alternatives. It is surprising, however, just how much we can learn from a simpler, yet intelligent, analysis and projection of the trends in the historical data, and thanks to BARB we have a readily available time-series of nearly 20 quarters of SVOD take-up data to work with.
The Figure below provides further technical details, and it is clear from the outset that neither a logarithmic nor an exponential trend projection provides any useful insights. The former, as well as being a relatively mediocre fit for the historical data, clearly underestimates the future potential of UK market SVOD take-up, while the latter, though a much better fit for the historical data, clearly generates unjustifiably optimistic projections. A simple linear trend, however, not only fits the historical data extremely well (predicting 99.3% of the variations in quarterly SVOD take-up between 2014-Q1 and 2018-Q3), but also appears to provide reasonable short-term forecasts, suggesting that if current SVOD growth trends continue for the next couple of years (and there are presently no compelling reasons to suggest they won’t) we could be looking at as many as 55% of UK households subscribing to at least one SVOD service by the end of 2020.
For longer term projections, however, a linear trend becomes problematic as it inevitably heads towards the 100% take-up mark, with SVOD take-up predicted to exceed 80% of UK households by the mid-2020s, and there are good reasons why this is unlikely. The quality of free-TV in the UK has always been a significant barrier to the number of households willing to pay (the compulsory BBC licence fee notwithstanding) for a TV service. In the US, for example, despite significant cord-cutting in recent years, around 78% of households are estimated to still have a pay-TV service, while in the UK only around 54% do. This resistance to paying for TV will (to some degree at least) almost certainly also extend to SVOD – why pay to watch Friends on Netflix (reportedly its most-streamed show in the UK), or even on Comedy Central for that matter, when you can currently also watch it for free on Channel 5? For many, a free service like Freeview Play, combining all the DTT channels with free catch-up/BVOD services (currently: BBC iPlayer, ITV Hub, All 4, Demand 5, UKTV Play, CBS Catchup Channels UK and Horror Bites), is simply enough, and broadcasters are investing significantly in the service to keep it competitive. It is also telling that in the US, where key SVOD services (notably Netflix, Amazon and Hulu) have been around for several years longer than in the UK, and where people are far more accustomed to paying for TV or TV-like services, SVOD take-up is currently estimated to have reached around 69% of households, and while there is certainly potential for further growth in the US market, it is difficult to envisage a realistic UK based scenario where SVOD take-up levels reach 80% or more by the middle of the next decade.
The question therefore remains, what are the likely upper limits of UK SVOD take-up and when are we likely to see the currently very significant rates of growth begin to level off. Clearly a more sophisticated approach than a simple linear projection is needed. Fortunately, there is an S-shape curve known as a logistic function that can help. It is often used to model the proliferation of new technologies through a population, beginning slowly with the early adopters, then spreading much more rapidly as it becomes mainstream, and finally slowing down again as the more reluctant late adopters are converted. The problem is that to model a logistic trend one first needs to stipulate a theoretical upper limit for the extent that one expects the technology to proliferate (the default being that it is taken up by 100% of the population), which is of course exactly what we are trying to find out – bit of a chicken and egg situation! We can, however, solve this dilemma by essentially doing some econometric reverse engineering and testing how well functions with different upper limits fit the historical data, and thus, through trial and error, arriving at the best-fit (i.e. the logistic trend with the highest R² value) solution. As it happens, the logistic trend that best fits the historical data (predicting 99.5% of the variations in quarterly SVOD take-up between 2014-Q1 and 2018-Q3) has a theoretical (asymptotic) upper-limit of 69% (i.e. the model suggests that the extreme upper limit of SVOD take-up in the UK is 69% of households). As can be seen in the Figure below, this results in the logistic trend model predicting that SVOD take-up will begin to progressively level-off in the early 2020s, and approach market saturation by the mid-2020s at around two-thirds of UK households. That being said, the results from any type of trend projection (even a relatively sophisticated one) need to be interpreted with a pinch of salt, as they rely on current and/or easily foreseen future market conditions applying over the forecast period (e.g. major new SVOD service launches in the near future, continued SVOD investment in original content, and the continued availability of attractive free alternatives), and cannot account for any significant disruptive paradigm shifts (e.g. as yet unforeseen disruptive technologies/services, the unexpected collapse of a key player, or major economic upheaval). Nevertheless, they do give us an idea of the likely direction in which we are heading, and the available data does suggest that a substantial minority (around a third) of UK households will remain resistant to SVOD.
So, while SVOD in the UK is likely to become even more mainstream, it will not be ubiquitous, which is good news for the UK’s TV broadcasters as this will go some way towards mitigating future losses in market share to the SVOD services. The last 5 years have certainly been testing times for schedule-based television, with worryingly significant year-on-year declines in younger audiences, and no indication that this is likely to level off for at least the next 2 years. That being said, and as I have repeatedly argued in my blogs this year, while one shouldn’t discount the fact that the industry is facing some very significant challenges, it is also very important not to underestimate the resilience and staying power of schedule-based television. It is easy to focus on the negatives and lose sight of the fact that schedule-based television remains something that we, on average, consume for several hours each and every day, and, despite unprecedented levels of market fragmentation, this also remains true for younger audiences. Ofcom’ s Media Nations 2018 (pp. 20-21) report puts the amount of time 16-34s spend watching video content across all devices (so this includes mobile device based viewing of video clips on YouTube for example) at 288 minutes per day, of which 46% (132 mins) is broadcast content [in the form of either Live TV (34%), Recoded playback (8%) or BVOD (4%)], while SVOD watched on a TV set only accounts for 10% (30 mins). In fact, schedule-based television remains the largest single video viewing category for 16-34s by a significant margin, while for the UK’s population overall, schedule-based television still accounts for a sizeable majority (71%) of video consumption, with TV set based SVOD accounting for only 6%.
Clearly there is a long way to go before schedule-based television is eclipsed, whether as a result of the SVOD services competing with TV on the big screen very directly, or the FAANGs (which of course also include SVOD’s Netflix and Amazon) competing for our screen time in general, or both. In fact, although there is no room for complacency, a far more likely outcome is that there will be convergence, with a new equilibrium where schedule-based television, though smaller, remains a major component of our daily video viewing routine. Binging on what you want when you want is great, but so is the communal Live TV viewing experience, as is catching up with shows that are still ongoing and being talked about, these are all things that meet fundamental needs in both the young and the old (albeit to different degrees). Schedule-based television and SVOD can co-exist and complement each other in the long-run, and the good news for us is that we’ll have even more great shows to watch on our bigger and better TV screens. To adapt a phrase from AMC’s Better Call Saul, the critically acclaimed Breaking Bad spin-off prequel, (which is currently available on Netflix in the UK, and where, in a very linear TV like manner, the latest season was released in weekly instalments the day after each new episode aired on AMC in the US): It’s all good TV man!
If you would like to receive the associated research notes to Dr Farid El-Husseini’s blog posts please email him directly on: firstname.lastname@example.org.