While some have been advocating switching off terrestrial television in the United Kingdom to save money, the cost of the online alternative has not really been counted. It may be convenient to consider the cost of online television as marginal or already accounted for, but the combination of additional household connectivity and broadcaster distribution could be almost a billion pounds a year to provide an internet-delivered equivalent to universal broadcast television.
The BBC has been among the leading voices in arguing that terrestrial television will be largely irrelevant beyond 2034. Given that the BBC is in the course of negotiating its future funding and purpose, arguing against the relevance of broadcasting might seem to some somewhat perverse.
One possible explanation is that the BBC is less keen on paying for its use of the transmission network and would prefer that this be subsidised separately by the government as national infrastructure.
An argument may be made that migrating to online delivery will deliver other public purposes by creating a connected nation that will support other policy objectives and enable wider use of digital services.
Those advocating switching off terrestrial transmissions suggest that the majority of the population can already receive online television and video.
That is demonstrably true in that around seven out of ten homes in the United Kingdom have access to an online video subscription service like Netflix.
The BBC heavily promotes its own iPlayer platform, which is now approaching 20 years old. It has now been around for a fifth of the history of the BBC.
The latest figures from the communications regulator Ofcom suggest that across all individuals the average usage of broadcaster online video services is less than half an hour a day. That compares to an hour and a half a day of television viewed at the time of transmission and a further 20 minutes a day of recorded playback.
Viewing varies by age group and availability to view. Among those aged over 55, the proportion of viewing at the time of transmission is much higher. It is undoubtedly falling, but online has yet to account for the majority of viewing of public service broadcasters for any age group.
Broadcast video on demand accounts for just 5% of all viewing. It represents about a quarter of all BBC viewing. That is after almost 20 years, with relentless promotion at breaks between programmes.
In other words, nearly three quarters of all BBC television programming is viewed at the time of transmission or is playback of a previously transmitted programme.
Even if some of that broadcast programming is delivered online, through platforms like Freely, it is clear that the majority of BBC programming is viewed at the time of transmission.
For the rest, the BBC is contesting the increasingly competitive space for viewing on demand. That does not seem to be a battle the BBC is winning.
The combined share of viewing of Netflix, Amazon Prime, and Disney+ is 15%, which is three times that of the BBC iPlayer. YouTube also has about three times the share of viewing of the BBC iPlayer. Those figures are as measured by Barb in the home through metering household routers in a representative panel of homes and do not include viewing outside the home.
That is not to suggest that the direction of travel is not towards more online and on-demand viewing or that there should not be a strategy for supporting that.
The challenge is how to deliver live television to the nation at moments that matter.
The football World Cup semi-final between England and Argentina was watched by a peak live audience of 21.5 million in the United Kingdom. Had England reached the final the peak audience would likely have been considerably larger.
The online delivery of live television for a country with a population the size of the United Kingdom would need to be dimensioned for around 30 million concurrent streams, or about one per household across the country.
That is actually easier to deliver if everybody is watching the same thing at the same time, like a World Cup final, but those are relatively rare events.
The ordinary scenario is that fewer homes are all watching at the same time, but they are watching programmes across a range of different providers.
That is entirely possible to engineer, but it does not come for nothing. The problem is that no-one seems to have published even approximate estimates of what that might cost.
Saving the cost of traditional transmission looks attractive if you do not consider the cost of the alternative. If the cost of online delivery is fully costed, traditional transmission may look more cost effective.
Of course, an online network is bidirectional and multipurpose. A large part of the cost of provision moves to the consumer, paid for through monthly connection charges.
Yet online video delivery is not free to media providers. With current approaches, broadcasters generally incur costs for every individual online stream delivered. This may seem reasonable while online is seen as incremental distribution, but when viewed as a substitute for the fixed cost of broadcast transmission it may compare unfavourably for large audiences.
For commercial broadcasters, the additional cost of online delivery may be partly justified by the potential premium
attributed to addressable advertising. If all video advertising becomes addressable that differential may be diluted.
For a broadcaster like the BBC, which generally does not carry advertising, all delivery is a cost of business, a cost of being in what might still be considered the broadcasting business.
The BBC spends about £600 million a year on content delivery across television, radio, and online, up by £67 million over the last year. So the BBC spends roughly one-sixth of its public service spending on distribution.
Alternative approaches to technical delivery online could considerably change the cost equation. Yet the vast majority of online video is currently delivered using unicast approaches that are now 20 years old and are arguably inefficient for very large simultaneous audiences.
These are all solvable problems but the proposition that existing online delivery infrastructure will support peak audiences of live programming at universal-service scale without substantial additional cost or engineering is questionable.
With about 28 million television households in the United Kingdom, assuming a monthly internet service provider connection cost of £30 per household gives an implied consumer cost for internet connectivity in the order of £10 billion a year.
Video accounts for around half of the data carried over fixed broadband networks. On that basis, a notional allocation of consumer expenditure on the connectivity infrastructure supporting online video is broadly around £5 billion a year.
Around 6% of households in the United Kingdom with a television do not currently have a fixed or household broadband connection. That is at least 1.5 million homes, representing over half a billion pounds a year in additional connectivity expenditure at an assumed £30 a month if an equivalent universal television service is to be delivered online.
People in the United Kingdom watch around 110 minutes of broadcast television a day, including live and recorded viewing. That amounts to about 45 billion viewer-hours per year. Assume that at large volume it costs roughly a penny per hour per viewer for a broadcaster to distribute video online, that is about £450 million a year. With its current share of video viewing, that would put the cost to the BBC at approaching £200 million a year.
While it is convenient to consider the cost of online television as marginal or already accounted for, the combination of additional household connectivity and broadcaster distribution could amount to approaching a billion pounds a year to provide an internet-delivered equivalent to universal broadcast television.
That does not include any additional costs for service delivery infrastructure, additional access network capacity, or customer equipment. A full analysis would be a much larger exercise, but it is safe to say that the cost of delivering television online is not nothing.