OTT Providers Must Address Delivery Network Access to Remain Competitive

Tuesday, September 18th, 2012
Screen Digest logo

Proponents of over-the-top (OTT) video delivery view streaming Internet providers like Netflix as the future of television, offering entertainment at a fraction of the cost of traditional pay-TV operators while also providing the potential for superior breadth in programming, ubiquity in platforms and technological advancement.

However, a detailed analysis reveals that the costs of OTT would rise to uncompetitive levels when such services are scaled up to meet the demands of large numbers of the average TV-viewing public. Because of this, platform-agnostic OTT providers like Netflix may have to change their business models as the market develops, perhaps investing in their own video content distribution infrastructures, similar to those already maintained by pay-TV companies. Yet as they become more like the pay-TV operators they are competing with, OTT firms will see their price advantage compared to cable, satellite and other delivery systems evaporate– and actually reverse–according to an IHS Screen Digest Insight Report from information and analytics provider IHS (NYSE: IHS).

“To serve the viewing needs of a mass-market audience, the content delivery network (CDN) costs for OTT streaming services would have to fall by a factor of as much as 25,000 just to reach parity with the most efficient broadcast technologies,” said Guy Bisson, research director for television at IHS. “At current prices, it would cost 1.2 billion euros in CDN costs alone for OTT unicast streaming to serve the population of the United Kingdom with the kind of high-definition (HD) viewing they are accustomed to. For the same price, 5,000 linear channels could be broadcast, nearly 10 times the number actually serving the U.K. today. When OTT unicast streaming services like Netflix are scaled up to suit the mass-audience television market, their advantages in cost, flexibility and technology turn into disadvantages.”

As presented in the figure below, the CDN cost of delivering satellite standard definition (SD) video broadcast remains stable as the number of viewers increases. However, the CDN cost-per-hour to provide OTT streaming standard-definition (SD) video rises as viewership grows, exceeding satellite when the audience reaches 8,000.

Simultaneous Views

“Even at just 8,000 simultaneous views, unicast streaming becomes less cost effective than broadcast–and this is a tiny amount compared to a typical primetime audience for linear TV,” Bisson said.

OTT Faces Efficiency Challenges in Large-Scale Deployments

With unicast streaming, it is as if every viewer is watching a completely different television channel at the same time. But even if everyone watches the same program and starts and stops at the same time, there are no efficiencies gained with unicast because 1,000 viewers still require 1,000 different streams. In fact, the more viewers there are, the more the cost increases.

In contrast, the efficiency of broadcast television comes from the fact that a single channel is broadcast once and reaches many viewers. Broadcast systems like terrestrial, satellite and cable are optimized for the simultaneous delivery of content to large audiences. For broadcast, it costs as much to reach one viewer as it does to reach 100,000 or 10 million.

The efficiency of broadcast is also its limitation in that only a restricted number of channels can be offered within the given bandwidth.

Conversely, unicast streaming’s advantage is that the number of programs or channels that can be offered is unlimited, because it makes no difference to the efficiency or cost if everyone watches a different channel simultaneously.

Practically speaking, however, OTT’s unlimited channels represent only a theoretical advantage, given that the cost of unicast video streaming climbs as it scales to larger audiences.

Alternative Business Models for OTT

Can OTT evolve then to meet the scale demands of the average TV viewer, or is it destined to be relegated to the role of delivering low-demand content?

One way OTT could be scaled up is if providers invest in their own infrastructure by effectively building their own distribution platform. This may not have to be a full-blown wired network, with all the capital spending demands of laying new fiber or coax. It could instead take the form of a halfway house of edge servers for localized content storage, creating efficiencies in CDN costs by delivering content once to each of a number of local hubs. In combination with a hybrid system that delivered linear channels using a different technology, this would go some way toward creating a platform-like service. But apart from the cost implications and inevitable knock-on effect on consumer-level pricing, this system would not give the OTT provider control of the linear component of its offer.

So, while scaling an OTT aggregator to meet the usage patterns of a primetime channel could be cost effective with suitable infrastructure investment, it does not seem feasible to scale to platform levels in order to match the service proposition of companies like Comcast without actually owning a platform.

What’s New?

Of course, there is nothing to stop OTT providers building out their own infrastructure.

The Google Fiber initiative in the United States is one example of this, but the most telling aspect about Google Fiber is the price point for its top-tier Internet plus TV service. At $120 a month, this offers no price advantage over standard U.S. cable TV and broadband services–and that’s with a considerably reduced channel offer. To be sure, a major portion of U.S. cable operator expense comes from affiliate agreements with key channels. So, to match cable’s TV offer, Google will have to increase its own content costs considerably, further impacting its competitive standing.

“If the future of OTT video streaming providers like Netflix lies in offering an infrastructure-based service that costs $120 a month, consumers doubtless will be asking ‘what’s new?'” Bisson said.