Streaming Content Cost Analysis: $1 Billion+ Per Percentage Point of US TV Consumption
Plus, Implications for Marketers, As Growing Spending On Ad-Free/Ad-Light Services Creates New Challenges
For anyone in London next Thursday March 22, I’ll be speaking at Video Week’s New Video Frontiers conference. See https://newvideofrontiers.com/ for more details. Hope to see you there!
Hypothesis: Content Spending Share Should = Audience Share
In the past I have performed analyses which suggested a positive relationship between spending on content and consumer time with that content, which in turn has conveyed to me the importance of assessing content spending in order to assess the relative strength of different video-based services around the world. I recognize that business models associated with streaming services may vary to some degree, that companies with streaming services may be at different stages of maturity, and that other factors may be at play as well, but in general I have presumed that the view that the cost to buy high-quality content should be roughly the same per unit of audience, regardless of which programmer or service wants to buy it. In streaming specifically, as different services look to commission programs, I would expect that competition for programming ideas that would drive subscribers and/or reduce churn should lead to comparable prices for a given unit of content, defined in terms of the audience it is implicitly or explicitly expected to deliver.
These analyses have implications for revenue and profit growth for the services themselves, of course, but because streaming does not lend itself to high ad loads for most services and because many consumers will choose ad-free offerings, the growth of streaming probably results in lower reach of ad campaigns and substantially more expensive campaigns over time.
New Data From Nielsen Plus New Company Data Drives The Analysis
To arrive at an updated answer I thought I would look at Nielsen’s new “The Gauge” data which covers all uses of television sets for the period of February 2023, and which came out on Thursday. The analysis here suggests that my expectations roughly hold.
To make comparisons between spending on programming and audiences, let’s start with numbers for which I have a high degree of confidence based on public disclosures of US-specific data. Here we can look at spending by the two biggest US-only services, Disney’s Hulu and Comcast’s Peacock.
Per Disney’s SEC filings, Hulu’s calendar year 2022 spending on all programming amounted to $7.8 billion. Of course, this includes the vMVPD Hulu Live which we want to exclude from the analysis to match viewing of the AVOD/SVOD service that’s tracked in the Gauge. Assuming programming costs for the 4.1-4.5 million subscribers on its vMVPD service last year was around $65 per month I can estimate programming costs for the SVOD service alone amounted to $4.5 billion, which resulted in 3.3% of all TV viewing during February 2023. 2022 viewing for the SVOD service appears roughly similar, so on that basis we could say that costs amount to around $1.4 billion per percentage point of viewing. Meanwhile, Comcast said at the beginning of 2022 that its content spend for Peacock would amount to $3 billion in 2022. While viewing was not broken out for Peacock on the Gauge until December 2022, the service saw 1.0% of all TV use in February 2023 and also for the prior two months.
For other companies, this analysis requires more guesswork. We know that Netflix accounted for 7.3% of total TV viewing in February 2023 per Nielsen, and figures for 2022 were similar on average. We know that global content spending (Netflix’ amortized content expense) was $14 billion during 2022. While the company doesn’t disclose country-specific content expenses, if we assume that the percentage of content costs allocatable to the US are slightly higher than the percentage of subscribers who are in the US (given the higher costs of US-based content and historical data conveying as much), I could roughly estimate costs of around $6 billion. That’s around $800 million per percentage point of viewing – much lower than the others, but perhaps unsurprising given Netflix’ scaled penetration and relatively longer history in many consumer homes.
Meanwhile, at Amazon, US streaming viewing on Prime Video amounted to 3.0% of all TV consumption in February 2023 and 2.7% on average during 2022. We know that global spending on video (including originals, live sports and licensed third-party content) amounted to $7 billion. However, once again we don’t know how much of this cost should be allocated to the US. If we assume it’s half, or $3.5 billion, that’s around $1.2 billion per percentage point of viewing if correct.
Finally, among those companies for whom we have useable data, Disney+ incurred $5.8 billion in global content spending during 2022, and again if we presume half, or $2.9 billion could be allocated to the US, I would compare this figure to Disney+’s 1.8% of total TV viewing during February 2023, which was slightly below the average of around 1.9% for all of 2022. This equates to around $1.5 billion per percentage point of viewing.
Other services such as HBO Max are harder to apply this analysis to directly, as the linear HBO service is still substantial with around 30 million subscribers who probably account for viewing that could be equal to or greater than the amount of viewing associated with the streaming product (which, incidentally, represented 1.3% of TV viewing in Feb. 2023). Moreover, programming costs are not broken out by the company on a country or product basis. Similarly, there is no data provided on viewing for Paramount+, let alone Showtime’s streaming product in the Gauge data, so a direct analysis of that service is not possible either.
Overall, for those services where we can make reasonable estimates, I calculate that content on streaming services costs roughly $1.2 billion per percentage point of audience. There are outliers such as Peacock, which is a newly scaling service and Netflix, which is much more mature, but the average appears to be broadly reflective of the business in the United States.
Total Content Spending By Distributors (Pay TV + Streaming) = ~$87 Billion in 2022
Building on this data, I can make some bigger industry-level observations. To look at the total amount of spending on professional streaming content, adding all streaming services up from this data and excluding YouTube’s non vMVPD product and its vMVPD product alike, 26% of TV consumption was associated with conventional streaming services during February 2023. On a comparable basis for the full year 2022, the equivalent figure was probably closer to 22%. Using the average cost figure of $1.2 billion per percentage point of viewing, this would equate to $27 billion of content spending for streaming services. Based on an analysis of programming expenses from traditional pay TV services, I can estimate that there was around $60 billion in spending on programming from conventional pay TV services, which occurred on approximately 60% of TV viewing covering both broadcast stations and networks and cable networks as well (recall that while broadcasting programming is technically available for free for consumers who use an antenna, most access it through the MVPDs who in turn pay broadcasters retransmission consent fees). If that’s correct, viewing of linear programming costs around $1 billion per percentage point of TV consumption.
Continued Spending Growth On Streaming Services = More Viewing on Streaming and More Challenges For Advertisers
Looking to some implications around the business in the future, I note that despite rhetoric around controlling costs, spending on streaming services is likely to go up while spending on linear services probably holds stable or declines. Increased spending on streaming is likely so long as there are large players with streaming services who lack legacy media businesses (Netflix, Amazon and Apple) who have reasons to continue increasing their investments, even if incrementally. For them, their increases in spending will likely contribute to share gains taken from anyone who cuts costs if everything else were held equal.
At the same time, some of the services have been explicit in anticipating spending increases, as we have seen with Peacock (Comcast has said they will increase spending on Peacock towards $5 billion in 2024 vs. $3 billion last year) and Paramount (whose spending should essentially double on Paramount+ from $3 billion to $6 billion at a global level). It’s true that Warner Bros. Discovery might try to curtail growth in spending, but their subscriber goals suggest that they’ll have to increase it regardless. Lastly, Disney has backed off of its subscriber goals in favor of a profitability focus, but similarly it seems unlikely that the company will be able to avoid spending more on its programming unless it wants to go backwards in size.
It's important to keep in mind that all of this has consequences for marketers so long as the bulk of streaming viewing is not likely to be ad-supported. This means that it will be increasingly difficult to cost effectively manage campaigns against reach-based goals. While connected TV-based inventory may help find those incremental points of reach in a cost effective way, most marketers who care about reach and frequency of video-based ad campaigns will either need to include YouTube (or TikTok, so long as it is around) in their plans or otherwise find other metrics or media to focus on.