BBH Report Home Page
May 23, 2002 Provided by System Dynamics Inc.

Cable's Conundrum: Set-tops and Moore's Law

Cable operators have a tough problem to unravel. On the one hand, they want to offer increasingly rich interactive services to their customers. Translate that to mean lots of software, graphics and processing to enable service delivery. Unlike satellite operators, their business model has always involved the MSO supplying (purchasing and leasing to the customer) the set-top. If an Echostar customer wants personal video recording (PVR) capability, no problem. The customer buys a set-top with a hard drive and appropriate software. If they want it, they pay (although the costs may be subsidized by the service charges on a committment contract). But if a cable operator wants to supply PVR functions, today's business model says the MSO has to make the capital investment in the box.

So imagine the MSO buys a box with PVR capability and other jazzy stuff. Instead of the box costing say $250, its price goes up to $450 or $500. Meanwhile, Moore's Law is inexorably at work, so by the time they've ordered, received and deployed some of the boxes, memory, processor, hard drive and other costs have dropped and the box is quickly outmoded, like last year's computer. In the face of such a conundrum, many operators decide to continue buying the lower-end boxes, which actually have a longer economic life.

Why not have the consumer buy the cable set-top? That's what some legislation and Open Cable were all about, right? Well, maybe. But first the MSO has to develop the channels and business model. And then the customer will go to the retail store, and is open to choosing not only what cable box they want but maybe instead choosing DBS. All of a sudden, sales incentives, side-by-side comparisons and lots of other factors have to be considered by the MSO.

The cable industry has seen this play out over the past few years as Motorola's DCT 5000 was the box that perennially was "almost ready"-- but constantly battled the issues of balancing costs, function, software etc. In the mean time, MSOs who had spent the money to deploy older cable set tops have been reluctant to replace them, if they can squeeze more useful life out of them. Rather than shelling out big bucks for new set-tops, software developers are cleverly shoe-horning more into existing boxes like Motorola's DCT-2000. For example, Motorola and Wink have integrated Wink's interactive TV service and Gemstar interactive program guides into the 2000, which are being deployed by Charter.

Add to the mix the fact that the average US home has 2.5 TVs (and may want the kids' and adults' TVs to both have access to advanced functions) and now we add in the problem and costs of whole home video, or investment in multiple set-tops.

Against this back-drop, vendors are providing some interesting solutions.

ICTV's HeadendWare

ICTV has been working for some time in the belief that, rather than chasing Moore's law in the set-top, MSOs should put most of the functionality at the head-end, with only a small footprint in the set-top. Although their solution originally involved their own hardware, their newest announcement of HeadendWare is a totally software-based platform that allows delivery of interactive applications and services from the headend. It uses a client-server architecture to enable "thin" digital set-tops to deliver the kinds of complex applications previously targeted for "thick clients".

Because their approach uses standard Intel-type processors at the head-end, the hardware costs constantly decline, directly benefitting from Moore's law. For example, today they specify 1.3 GHz Pentium 4 motherboards, but will be moving soon to 2.3 GHz following the processor cost curves. Their modules are interconnected with Gigabit Ethernet, so that the MSO can organize modules based on their particular plant architecture. ICTV has designed their own software based multiplexers, which are the only things that need to be in the headend and nodes. They indicate that a single rack can serve 500,000 ICTV subscribers, important in the tight spaces often available in cable locations.

ICTV has been honing their approach for some time now, and has substantially reduced their capital costs. In a phone interview, ICTV CEO Wes Hoffman told us that their first-time capital cost now comes out to $4 per digital subscriber, including head-end hardware, up-converters, software licenses and the first year of support.

ICTV is now concentrating on evangelizing application developers. Because it is based on Windows, HeadendWare provides developers a familiar way to develop and deliver applications to all the thin clients out there. There is nothing proprietary and different that application developers need to know or do -- Wes told us the (slightly tongue-in-cheek) three elements of their content design tool kit: 1. Don't use "hot red." 2. Design for TV. 3. Good luck! Interested application developers can get information from Anna Tarnay, .

HeadendWare is installed at Charter in Kalamazoo, Michigan; ICTV is working with Charter and Digeo on commercial deployment. They say they are also working with other well-known but unannounced MSOs on trial deployments. Since all elements of their approach are ready and available today, we'll be watching how the market votes with their dollars.

[Editor's note: We are on ICTV's Advisory Board and have consulted for them.]

( )

Ucentric's Whole-Home Entertainment Solution

Ucentric's key story is also about using software to generate new revenue streams for MSOs. Other key concepts make lots of sense to us: client-server architecture and a notion of low cost (thin) clients for all but the primary TV.

We interviewed Paula Giancola, Ucentric's VP of Marketing, to find out what was getting cable operator's attention at the recent NCTA Show. The big competitive issue facing cable operators today is clearly satellite services. At year-end 2001 Echostar and DirecTV together accounted for about 17.5 million subscribers and their numbers keep growing. Satellite providers' ability to provide new features like PVR more quickly than cable has been a real competitive threat. As we saw in the ICTV story above, providing thick clients with PVR capability is economically challenging - and especially so for multiple TVs.

Ucentric's software platform provides multi-TV PVR functionality from a single box. It is designed to be integrated with boxes from existing providers; for example, it has been integrated into Paceís "digital home gateway" (set-top). It stores content and distributes digital entertainment and communications services to the primary TV and stereo, and to computers connected over a wired or wireless network. Additional TVs receive the services through small low-cost boxes ("media clients"). Services include multi-TV personal video recorder functionality (PVR), multi-TV Video on Demand (VOD), multi-speaker digital music, and shared Internet access, throughout the home over a wired or wireless network.

Ucentric's server software includes a Linux-based operating system with Quality of Service (QoS) enhancements, an applications services engine, a customizable user interface and a suite of media-rich applications. Ucentric has been through six trials, including four with MSO partners, and has incorporated feedback from both the service providers and the 150 user homes where it was trialed. They are now in their third generation of software.

Mindful of our car radio story and Cooper's question of what you get when you cross a computer with a TV (see above), we asked Paula what they had learned about the user interface from customer reactions. She said they have refined their user interface, making it less "mouse like" and more readily used with the TV remote, more readable on the TV screen and more in line with customer expectations about how to interact with the TV. They have teamed with AGENCY.COM Ltd. which is helping to develop the user interface design for its multi-TV personal video recorder (PVR) application. Ucentric is incorporating AGENCY.COMís design templates and style guidelines into the TV-based user interface for the entire suite of whole-home entertainment applications. We haven't had the opportunity to play with it yet, although Ucentric did demonstrate it to some potential clients in their private suite at NCTA. Hopefully they are meeting their design goal of creating a "sensible and simple means to manage and control their digital entertainment".

The jury is still out on whether cable operators will adopt Ucentric's platform or any that departs from the tradional "one box to a TV" model. Now that they've partnered with a major manufacturer, and will have software for deployment later this year, we'll watch to see Ucentric announce MSO rollouts.

( ) ( ) ( )

(For our previous article on Ucentric see BBHR 11/14/2001.)