Member Mailing #264: Subscribers vs. Viewers in Sports Streaming

Why bundling has its limitations in streaming, especially in sports streaming

There has been an emerging, evolving discussion of the growing role of bundles in the streaming marketplace. Some of the more recent chatter has been driven by the merger of WarnerMedia and Discovery: that merger will result in a bundle of WarnerMedia and Discovery channels and streaming services. The outcome also reflects the shortcomings of AT&T’s bundling strategy relative to its competitors Verizon and T-Mobile.

The more familiar definition of a bundle comes from the cable model: MVPDs aggregated audiences at scale through fiber or satellite connections, and bundles made the economics of consuming linear cable more affordable for those audiences. In that model, both demand and supply are effectively guaranteed by the infrastructure of cable distribution. It was best explained by the economics 101 definition of a natural monopoly: MVPDs had high start-up costs but offered powerful economies of scale. The bundle reflected the pricing power of MVPDs resulting from that scale, allowing them to effectively trade the widest possible access to consumers in exchange for lower retransmission fees.

With the ongoing loss of cable households (now at ~85MM), there are fewer subscribers to justify lower user costs, making the rise of streaming as cable’s replacement increasingly inevitable.

But, in streaming, the objective for every service has been to disrupt that bundle model by circumventing MVPDs and building direct-to-consumer models at scale. Netflix has found success reaching as many as 68MM American subscribers directly, and both STARZ and AMC+ have found successful streaming models at 10-15% of that scale. Otherwise, without the infrastructure of cable, no streaming service has aggregated broadband households across the U.S. with a business model that mirrors the infrastructure that led to over 110MM households with cable. Without hardware and distribution infrastructure to ensure both scale and stability in its subscriber base, the DTC relationship in streaming is ephemeral.

That limits bundling as a solution in two ways. First, bundling may aggregate subscriber audiences at scale, but that relationship is ephemeral. Secondly, personalized user experiences create a higher probability of viewing audiences at scale than bundling. Two perfect examples of this are Hulu and Amazon Prime Video.

“The New Bundle Package”

Strategic media consultant and former senior Fox executive Patrick Crakes has argued that the legacy cable bundle may be dying, but bundling is not dead:

the route to scale, pricing power and subscriber stability for streaming runs right back through the business strategy that built the legacy pay-TV system: bundling.

Bundling “allows scaled content offerings and creates high switching costs across multiple fronts, which in turn results in long-term stability.” As streaming services increasingly replace cable bundles for video consumers, they ultimately will end up needing bundles with the big three strategic economic components - scale, pricing and stability - that make up the core moat of MVPD bundles. Crakes writes:

Streaming platforms face the double whammy of exponentially increasing content creation and acquisition costs combined with annual technology investments that can never be skipped or delayed. These large capital expenditures mean the cost component of the profitability equation is always accelerating, making profitability via cost controls and delayed infrastructure investments hard to achieve. 

This means the revenue side of the house has to grow at a brisk pace, but the challenge there is growing subs while maintaining structural price integrity. While currently there’s probably some room for higher prices given the still-low monthly rates, limited content offerings at some point force the realities of price elasticity to kick in, ending organic sub growth or, even worse, driving sub losses.

In other words, without cable’s distribution infrastructure streaming platforms face a difficult future scaling and reaching viewers. The operational expenses of running a streaming business are going to continue to grow. Streaming services will be limited in their ability to raise prices due to the price sensitivities of the target streaming consumer. Bundling is an effective, proven means of keeping those prices down, and attracting subscribers.

But, we will not see “a recreation of those video-dependent established bundles” of MVPDs. Rather, Crakes predicts we will see “a new type of digital bundle that many established and new content owners and video distributors are experimenting with”, a bundle “with enough content and services that there’s a clear value proposition for the consumer at higher prices.”

That is a bearish take on video streaming, but a bullish prediction for the likes of Disney and Amazon. One version of that bundle looks something like Apple’s One Bundle (which I wrote about in November) of TV+, music, news, and games. A more relevant version looks like Amazon’s Prime bundle of services (shopping benefits, free shipping, books, video, music).

Another relevant version of the bundle is the Disney-Verizon partnership, which Crakes argues “is a precursor of new bundles that will have many more types of services and content within them”:

Allowing Verizon customers to get access to Disney+ for free not only drives sampling but also provides valuable opportunities to push the entire Disney content bundle organically. For Verizon, the added content from Disney+ makes their broadband and communication services more valuable at no cost to customers.

I like this Disney-Verizon example because it is a great example of how bundling helps to drive scale for both companies. 20% of sign-ups for Disney+ in Q4 2019 (5.3MM) came from the Verizon bundle offering. Assuming that has held true through Q1 2021, and Disney+ is now at around 40MM households in the U.S., 8MM Disney+ subscribers came from Verizon.

In Q1 Verizon told investors ”more than two-thirds” of those customers “have maintained their subscription, either through their Verizon direct billing relationship or by opting into one of our newest Mix & Match plans with the Disney bundle [of Disney+, Hulu and ESPN+] included.” That means at least 6MM customers, or 15% of Disney+’s installed base, have come from the Verizon bundle.

In other words, Disney+ achieved 80% of its existing subscriber base without the Verizon bundle. It is reasonable to assume that Hulu also would be at 80% of its current scale its SVOD has (37.8MM), or 30.2MM SVOD subscribers, without the Verizon bundle. That would put it about halfway between the numbers it reported in fiscal Q2 2020 (28.8MM) and in fiscal Q3 2020 (32.1MM).

So, the Verizon bundle is evidence that “a new type of digital bundle” can indeed drive subscriber audiences at scale, and at a scale that makes a substantive difference.

But, these are subscribers and not viewers. The Verizon bundle is not concerned with subscribers as an outcome, as Verizon is rewarded with per subscriber payments (per month) from Disney.

This leads to the logical question: can “a new type of digital bundle” drive viewing audiences at scale?

Subscribers vs. Viewers

I think this distinction is important because it reflects the business logic of the direct-to-consumer conversion funnel. Effectively, bundles accomplish the business goal of growing subscribers, but do not always accomplish the goal of driving viewership. That means there is an additional burden, if not marginal operational expense, incurred by a streaming service.

Sports streaming is a perfect example of this challenge: Amazon Prime has a $1B/year deal with the NFL for Thursday Night Football, and Disney has $400MM+/year deal with the NHL to stream 75 NHL games per year on ESPN+ and Hulu. Both of those deals assume, and therefore require, viewing audiences at scale.

The fourth quarter marked Prime Video’s strongest viewership for live sports globally. In the U.S., Prime Video’s exclusive coverage of the San Francisco 49ers vs. Arizona Cardinals game on December 26 drew an estimated 11.2 million total viewers and delivered the highest digital average-minute-audience ever for an NFL regular season game. In the UK, the number of customers tuning into live Premier League football grew for the second season as millions watched 22 live and exclusive matches on Prime Video. In addition, millions of Prime members streamed live, international rugby for the first-ever Autumn Nations Cup tournament; and in India, Prime Video announced its first foray into live sports, with the acquisition of India territory rights for New Zealand Cricket through 2025-26.

In other words, Amazon can consistently drive Prime subscriptions at scale, but it cannot yet consistently drive Prime sports viewership at scale. Even its Thursday Night Football figure of digital average-minute-audience if 4.8MM is less than one-third the NFL’s average audience of 15.4MM viewers (live + same day).

The same is true for Hulu, which has grown by over 50% since its merger with Disney. But, the stakes are different for Hulu: ESPN+ has only been integrated into Hulu for less than 90 days. So if Amazon is still in the early days of figuring out its sports streaming strategy for driving viewership, Disney has only just begun to test what will drive sports streaming viewership.

However, for Hulu, it has a fraction of the challenge for NHL viewership: during the regular season, NHL games averaged 391,000 viewers across NBC and NBCSN; and, the 15-game Wednesday Night Hockey package averaged 508,000. Basically, Hulu only needs 1% of its subscriber base and 3.6% of ESPN+’s subscriber base (13.8MM) to recreate NBCU’s audiences digitally.

In this light, Disney and Hulu may be better positioned for driving streaming viewership than Amazon in sports streaming.

Conclusion

The distinction of subscriber audiences at scale from viewing audiences at scale is a necessary one in sports streaming. The linear bundle all but ensures pricing efficiencies that guarantee both subscriber and viewing audiences.

But, bundling streaming services is not a solution for driving viewership at scale: the key variables that ensure scale for sports viewership in the cable bundle - pricing discounts, a stable subscriber base at scale - mean little to sports streaming viewers.

This reflects two problems for streamers with sports rights.

The first is a marketing problem: sports viewership will difficult to scale in streaming. Amazon and Disney are unique because they are able to rely on their respective ecosystems to drive scale in viewership for sports streaming on Amazon Prime Video and Hulu, respectively. The question is whether that scale will be enough.

The second is an advertising problem, one which I wrote about before in “Member Mailing #256: Amazon, NBCU, The NFL, & Addressable Advertising”.

…Amazon’s narrower, CTV/OTT streaming-focused value proposition means an advertiser will be able to buy CTV/OTT inventory elsewhere within the Amazon ecosystem, and not just a portfolio of different types of inventory.

Ad buyers are confronted with a choice between buying OTT/CTV ad targeting with contextual linear inventory, first, and CTV/OTT ad targeting that may be more valuable than linear TV, second. Hulu offers the former, and Amazon Advertising offers the latter. But, both will need viewers at scale to keep advertisers happy.

Amazon’s and Hulu’s shared advantage is that their respective bundles can create a stable, existing subscriber base at scale. Their shared problem is that there is no guarantee of viewership from their respective subscriber bases, or from subscribers obtained via their respective versions of a “new type of digital bundle”.

It is a problem with no obvious answers, not even from a “new type of digital bundle” that will emerge in the linear era.