Complexities of Cable PCTV Agenda Portend Disruption on Many Fronts

Jeffrey Bewkes, chairman & CEO, Time Warner

Jeffrey Bewkes, chairman & CEO, Time Warner

April 13, 2009 – What began as the cable industry’s attempt to mount a coordinated response to growing consumer demand for access to TV programming on the PC has morphed into a complex open-ended agenda with major implications for business models and technology platforms alike.

Public statements by senior MSO and programming executives at the industry’s recent National Show in Washington, DC revealed there is broad concurrence on the need to create a new approach to providing content online that satisfies consumers while protecting existing revenue streams. But unpublicized details of the uncertainties surrounding approaches to advertising, content distribution, security and pricing strategies suggested it could take longer to achieve a cable TV-to-PC service than many players expect.

For example, a senior industry executive confided there may well be three areas of technology advancement that will require new initiatives on the part of CableLabs, including a new version of the broadband DOCSIS (Data Over Cable Service Interface Specifications) standard (see accompanying story). New approaches to content protection and aggregation might also be on the CableLabs docket before long, he said.

In general, programmers appearing at the National Show voiced support for the idea of offering their content to cable subscribers as a more complete reflection of their full lineups than they offer on the Web today. Acknowledging their experiments with Internet content are works in progress that have not generated significant new revenues, they made clear their desire to protect existing TV-centric revenue models, including the $25-billion revenue stream generated through cable subscription fees

Comcast, with its OnDemand Online initiative, and Time Warner, pushing TV Everywhere from the programming side, are spearheading negotiations among cable operators and programmers to determine what the business and operations models would be for such services. At the National Show Time Warner chairman and CEO Jeffrey Bewkes announced Time Warner was forging ahead with launch of HBO Go, which HBO subscribers will be able to access through their cable providers’ Web sites.

“We’re all being too slow,” Bewkes said. “We should put up all our networks on the Internet, out on broadband right now, get it out on home screens, broadband screens, put it on the Hulus and YouTubes, but only if people are subscribing to the video plant.” Comcast chairman and CEO Brian Roberts echoed the theme at another session, contending that, for programmers, a cooperative arrangement with cable on broadband distribution of content represents “a new opportunity to try to monetize in this horrific advertising environment.”

News Corp. chairman Rupert Murdoch and Viacom CEO Philippe Dauman, appearing on a panel with Bewkes, concurred with him on the need to do a better job monetizing Web-delivered content. “Nobody is making any real money from the Web except search,” Murdoch said. “We have to monetize it.” Dauman, while noting that agreement on a cable TV-to-PC business model was a greater challenge for ad-supported cable networks than it was for pay TV services like HBO and Showtime, said he thought operators and programmers could find a “middle ground” that would protect both sides’ interests.

Some press reports suggested Robert Iger, CEO of The Walt Disney Co. voiced opposition to the TV-to-PC monetization concept when he said that “preventing people from watching any shows online unless they subscriber to some multichannel service…would be something we would find difficult to embrace.” But, in fact, Iger was only stating the obvious, which is that programmers will not endorse a complete ban on free access to selective offerings of programming such as Disney provides through its various online outlets.

When it comes to full access to the complete programming lineup, which is what the discussion is really about, Iger was as supportive as others. “With authentication (of subscribers) in place, streaming full networks online would be an interesting and potentially compelling feature for consumers, and we are certainly open to exploring that possibility,” he said.

How to accommodate authentication is one of many issues that MSOs and programmers are working on in hopes of making services like OnDemand Online a reality before the year is out, said Marty Roberts, vice president of marketing at thePlatform, Comcast’s Web publishing arm. As the supplier of portal support services for Comcast, Time Warner Cable, Cox Communications and Cablevision as well as Web publishing services for many programmers, the company is playing a key role in working through various operational challenges for the industry as a whole, Roberts noted in an interview.

“No one is locked down on a strategy,” Roberts said, noting that decisions remain to be made on whether the broadband content will be offered free as part of cable TV subscriptions or at a premium and whether people will be given the opportunity to subscribe to a broadband-only cable TV programming package. While MSOs and programmers are negotiating with each other on a one-to-one basis with regard to licensing, usage policies, advertising and all the other core business issues, Roberts said, they are coordinating strategies as much as possible, especially when it comes to technical issues.

“The whole story leaped into public view six months earlier than we expected,” he said. “So there’s still a lot to work out in the way of business models, technology issues, licensing models, other things.”

But thePlatform has been on the case with Comcast for some time, which gives everyone a head start as they confront the many issues. “We’re discussing a lot of the underpinnings that we’ve been working on for OnDemand Online with the rest of the industry as well,” Roberts said. “The goal is to achieve a broad solution that will work across multiple MSOs and programmers.”

It’s a complicated task, made more so by the fact that the goal is to offer programming in a la carte on-demand mode rather than simply to stream it linearly as a replication of the cable TV channels. “Mostly we’re talking about non-linear distribution,” Roberts said. “It’s yet to be determined whether programming will be offered live [as broadcast] as well.”

On-demand availability of virtually all of cable TV programming over broadband would replicate the Web user experience, thereby meeting consumer expectations for entertainment in the Internet domain. But it would bring with it new requirements for the supporting distribution infrastructure and for advertising, where dynamic ad placements replacing the original live broadcast ad placements would be required for monetization in the new Web-centric mode.

“From an advertising perspective, linearly delivered programming would be easier to deal with, because you could leverage the existing transport infrastructure,” said Joe Matarese, vice president and general manager of on-demand strategies at Arris. “On demand requires a different infrastructure for making ads available and for placing them.”

Media delivery servers must have “multi-protocol capabilities on a converged delivery platform,” Matarese said. And it “probably makes no sense to use MPEG-2 transport,” owing to the need for descriptors to support new modes of ad placement. While the SCTE 130 standard, which sets specifications for identifying what ads to place and where to place them, is “completely leverageable,” other aspects of the advertising architecture for this new service “are likely to be different from the protocols we’ve been looking at.”

But, from the perspective of the broadband world, such capabilities are now a given for advertising, Roberts noted. “The ideal in terms of addressable advertising, telescoping, interactivity is already happening in broadband video today,” he said. “It’s second nature to us. Our job is to explain to the MSOs how this works and to explore with them what the optimal strategy will be.

“One of the advantages of our system is that you can set up different types of ad policies to a pretty complex degree,” he continued. “For example, you can have a policy that says 80 percent of the ad space in a given program gets filled by one entity and the other 20 percent by another, and if the party with the 80 percent control doesn’t have an ad to deliver, the request goes out to the other party’s server.”

“Nothing has been decided” with respect to advertising business models, Roberts added. “At thePlatform we’re thinking ahead to handle situations where there are different sources of ads with various failover scenarios.”

Indeed, Roberts noted, its thePlatform’s job to be thinking ahead on several fronts even though so much remains to be determined at the negotiations level. “We have a good working solution that we could deploy tomorrow,” he said, “but it’s important to over time that we build a better foundation to how it works.”

Authentication and interfaces to billing systems are already a component of the digital store module which thePlatform supplies to process orders for pay-per-view content offered by InDemand through Comcast’s Fancast and Time Warner Cable’s RoadRunner portals. Enhancements to those processes are a big focus of the new initiative.

Here again uncertainties abound on basic strategy. For example, some operators want to confine access to programming through their own portals while others want to follow a syndication model that allows people to get limited-access content from programmers’ sites as well. The latter case would require an authentication model analogous to the federated ID strategy employed in social networking under the OpenSocial and other models (see January ScreenPlays, pp. 1 %26 36), where the same identity code is used wherever someone needs to be authenticated for access to content.

Some operators are even talking about promotional hooks, such as offering a free weekend with a coupon. And everyone is trying to figure out how to accommodate subscribers’ access to content when they’re beyond their cable operator’s footprint. If access to restricted content requires authentication based on the modem ID every time a user requests a show in order to ensure the user isn’t just sharing their ID with friends and neighbors, it would be difficult to provide a service that is literally “TV Everywhere.” So some thought is being given to requiring just a periodic authentication of the user with the modem on the assumption that, if the user has been accessing content on the modem and then isn’t on the modem the person is probably traveling and should be trusted based on recent usage patterns.

“Whatever the model turns out to be, it’s just an evolution of our existing system,” Roberts said.

Another area of uncertainty concerns how content will be secured and what the means of conveying usage policies will be. “We’ll start with stream encryption and then add DRM (digital rights management) over time,” Roberts said. Historically, he added, the only DRM thePlatform supported was the one used with Microsoft’s Windows Media, but now there’s a need to embrace next-generation systems that facilitate policy enforcement and protection across multiple devices.

For example, the platform adopted by the Marlin consortium provides a Common Domain approach to content protection, where encryption keys can be moved about among any number of devices assigned to a user’s domain. In contrast, the protection mode known as Digital Transmission Content Protection over IP (DTCP-IP) was designed to limit authorization of usage a particular piece of content on just one device beyond the PC in order to assure content owners they would be able to generate new revenues with multiple device usage.

CableLabs, which has endorsed DTCP-IP in consort with the home networking consortium known as DLNA (Digital Living Network Alliance), may become involved in defining a new protection model for the TV-to-PC service, possibly by proposing extensions to DTCP-IP, noted the executive cited earlier, who spoke on background. However, he said, changing the standard to suit cable operators’ agendas or adopting another such as Marlin might prove problematic with programmers.

Another big area of concern in the shaping of the industry’s new TV-to-PC agenda is the degree of coordination that will be required in establishing a supporting network infrastructure for the aggregation and distribution of content in high-quality mode that avoids the quality issues imposed by the public Internet. “This is another area for CableLabs to look into,” the executive said. “You need a centralized point for storage and a means of distribution to local storage that maximizes efficiency and quality. There are a lot of technical details that have to be agreed to as well as business aspects in terms of how costs are shared.”

Indeed, Roberts noted, without reference to any role for CableLabs, the “four 9s” performance that thePlatform has always maintained in supporting Web-based content publishing and distribution is “nowhere good enough” for what the cable industry has in mind. “We need to carrier class,” he said. “Video on demand has to stay up with carrier class reliability and economics to make sure ads always show up and to deliver the experience people are accustomed to with cable programming.”

As for the local access side of the network, the potential demand for a high volume of video content raises question about the utility of the cable industry’s local broadband architecture as its currently constituted. Even the new DOCSIS 3.0 platform with channel bonding and the choice of what is known as I-CMTS (Integrated Cable Modem Termination System) and M-CMTS (Modular CMTS) for positioning of RF QAM (quadrature amplitude modulation) capabilities inside or external to the CMTS does not fully address cost issues associated with delivering a high volume of IP-based cable programming, the senior industry executive noted.

“The CMTS was not designed to transmit what amounts to an IPTV programming service,” he said. “Some people are talking about leveraging the modular option, bypassing the CMTS altogether by encapsulating IP content in DOCSIS 3.0 frames at the edge QAM. Others are talking about changes in the architecture of the CMTS itself. This could become a project for CableLabs.”

The scope and complexities of the ambitious TV-to-PC undertaking suggest there is much more to the agenda than leaders have discussed so far – possibly more than they’ve yet considered. If the goals, including new standards and centralized on-demand distribution architecture for IP-based programming, are met, will the industry be content to operate this service as a separate domain parallel to and replicating much of the next-generation time-shifted and advertising capabilities now being developed for the existing cable TV MPEG-2 based infrastructure?

“It’s hard to imagine creating something this immense and powerful and then trying to do the same thing in the TV realm,” said another senior industry executive, speaking on background. “What may really happen is you’ve got an immediate demand-driven need for an IPTV architecture that everyone wants to put in place as quickly as possible. Once that’s done, everyone is probably going to stand back, look at what they’ve built and realize they now have the next-generation service architecture they need to move ahead on the TV front. A lot of people don’t want to admit it, but this is our way of migrating to IPTV.”