Hildenbrand breaks the mold
No digital hand-wringing over at Cablevision Systems: Its plans to initiate a 100 percent swap-out of existing analog and digital boxes make it the most bullish MSO in the land. It happened last year, on Sept. 16, when Cablevision handed to Sony a $1 billion order for 3 million digital set-tops. Not only did Cablevision raise the bar on how to attract digital subs to cable—it also leapfrogged, by several tall jumps, existing digital set-top muscle. Foolishly risky, or smartly visionary? Probably the latter: Cablevision owns historical status as the MSO with the highest revenue-per-subscriber. Most of the technological foresight behind Cablevision's maneuvers is coming from Wilt Hildenbrand, executive vice president of technology and engineering for the MSO. Leslie Ellis, an independent analyst and author, caught up with Wilt recently for a digital status report. An edited transcript follows.
CED: Instead of approaching digital TV like your MSO peers—"Hi, would you like to pay me $10 more and get a digital box?"—you're saying, "Hi, here's your new box, hope you like it." You're swimming upstream here, as compared to your MSO peers. What drove that strategy?
Hildenbrand: It isn't as risky as it appears. It's actually reasonably straightforward. The plan was made up of a few major components. First, we didn't want to be bandwidth or technology constrained, so we have worked on completing the network buildout, and second, we didn't want our customers to be left wanting for channels. By doing that, we wouldn't be eroded by DBS, or other competitors for that matter.
CED: Is DBS churn an issue?
Hildenbrand: It hasn't been, and it's not now. Over the last few years, while we kept building out for new services, we were accused of not having a digital strategy. But actually, our strategy was to build out the network, with all the interactive capabilities that we've spoken about—HFC, 500-home nodes, etc.—and make sure everyone had at least 77 to 80 channels while we waited for the right digital product. Well, guess what? Here it is, the right digital box to unlock our strategy. At the same time, our build-out, by the end of 2001, is something like 90 percent complete. So here we are with these networks capable of doing high-speed data, telephony and digital video. The thing is, we didn't want to put in a box that was, basically, a channel expander. By that I mean, the early (Motorola) DCT-2000s, when we looked at their price against their capabilities, didn't seem to make any sense for us. We wanted something that would really leverage, on the TV set, as well as in the home in general, all of the services our network was capable of delivering.
CED: Such as?
Hildenbrand: We'd done two VOD trials. One of my favorites was the one with AT&T (in the mid-1990s), and the other was with Diva in New Jersey. Through that we recognized that VOD dramatically changed how pay-per-view ran as a business. Also security (is) no small matter (and) staying ahead of service theft gets harder and harder: The weapons get better and better, so we wanted a system with renewable security. Plus, within some reasonable timeframe and service penetration, it costs as much in hard capital and staff training to build the digital backend to support 10 percent of the base as it does to do 100 percent of the base. So the thinking was, if you only make it available to 10 percent of the base, you're also only getting 10 percent of the potential revenue and you miss the chance to generally upgrade service security. Lastly was simple bandwidth recapturing: If you can deliver much of the product digitally, in addition to what was said just before, you can reclaim more of the former analog bandwidth for VOD growth, high-speed data growth, HDTV, and whatever else comes along.
CED: It's an all-digital box, then? No analog tuners or descramblers?
Hildenbrand: No analog descramblers. It has an analog tuner. All of these decisions start to act like dominoes: Once we had the bandwidth, we could get rid of analog scrambling—since that requires duplication of services for a time. Then it was no longer required to only look at Motorola and S-A products. We could look for a company that knew consumer electronics and customer relationships. That's how we wound up picking Sony, and focusing on services beyond the box—that's what Sony does for a living. With Motorola and S-A, once they've sold you a box, the next thing they have to do is sell you the next one.
CED: You have specified two boxes—a primary and a basic. Why two?
Hildenbrand: It's an outgrowth of the 100 percent change-out model. You can't afford to replace every single box with a $400 one, for no additional charge. I'm biting my tongue here, but in a sense we did need a simple channel expander, as well—a basic, if you will. But, since the primary box doesn't support analog descrambling, we had to design a second, less feature-rich unit. So we have a $250-ish box, for all the extra outlets in the house. And we have a $350-ish box for the primary outlets. We wound up with two because of the simple cost of giving away, for no additional charge, that much capability.
CED: Run through what's in the primary box—what kind of processor, how much memory?
Hildenbrand: The primary has a 300 MIPS processor. We've doubled the memory since we last talked, to make stuff load faster. It'll have 32 megs of RAM, 16 megs of flash, 8 or 16 megs of video D-RAM. It has a DOCSIS and a DAVIC modem, a 3D graphics card, a tuner, a conditional access card slot, and a smart card slot, for an affinity or money card. Three USB ports—two masters and one slave—two 1394 or I-link ports. Out the back will be digital audio, S-video, NTSC, RF in and out.
CED: Geez. It slices, it dices, it julienne fries. What's in the basic?
Hildenbrand: That one has a 100 MIPS processor, half the memory of the primary, no DOCSIS modem, none of the other USB or 1394 connections, no money-card slot. The user interface is also all HTML. So it's all based on connections to a server.
CED: Let's talk about the software. What about the operating system and middleware?
Hildenbrand: The operating system is Vxworks, from WindRiver. Sony's designing and building the middleware. The browser is from Spyglass.
CED: How did you draw the boundaries around what the operating system does, versus the middleware and the browser?
Hildenbrand: I can't give you the diagram. Suffice it to say, it's like breathing. It's run by the brain. Everything else is optional.
CED: Are you following what's been happening with OpenCable? The middleware has been specified as an executable engine, that Sun will write, and an HTML presentation engine, that Liberate and Microsoft will write.
Hildenbrand: I haven't paid specific attention to that aspect.I think it's fine as long as everyone follows it. We've been almost evangelical—maniacal, even—about "open" and open standards in this thing. So at that level, if everyone does what he or she says they're going to do, that works for me. If everybody publishes rules, you can manage. Look at it this way: TV sets are built to an open standard, but no one TV is ever the same as the next. You'll have similarity. Having a baseline is the important part. As for OpenCable, if we all sat back and waited for every potential aspect of it to be completely defined, then none of us would be able to move forward at all.
CED: So, how about a status report. How's the rollout?
Hildenbrand: Status report is, I've got production hardware and software coming in-house next Friday. (Editor's note: Hildenbrand was referring to Friday, Oct. 6.) That's when we start in-house test and integration. Assuming that goes well, we go to beta in the beginning of November.
CED: Beta is with "friendlies?"
Hildenbrand: Friendlies, yes. Employees. Assuming the friendlies remain friendly, which they should, this is our second beta; (the) first consumer launch happens in December. No, it's not the best month in the world to put new services in front of customers, but everyone, including Sony, has been working very hard to get to this point so we can at least get the pilot—that is, first consumer test—started. We'll be ready.
CED: You've slipped some.
Hildenbrand:If you do the math, how much have I slipped? About six or seven months from where we started. From a clean sheet of paper, that's not bad.
CED: About two years of not bad, if you compare yourself to some of the other MSOs, their vendors, and the start-to-finish timeframe for digital video.
Hildenbrand: Right. So, is there another month or so of slip in this thing? Probably. We'll find that out from the pilot testing with consumers. But that's OK. We're looking pretty decent. Sony is a good partner. They're incredibly technologically adept.
CED: Do you view the Sony box as the gateway for in-home networking, telephony, high-speed data?
Hildenbrand: Yes on in-home networking and phone. High-speed data is one of those tricky ones. For connecting on PCs, we've already got the cable modem thing going full blast. This whole thing is a way to put—I hate to say Web pages—but rather Internet-style technology, on TV, mainly to enhance the television experience. Because Internet content looks like hell on TV, and was really designed for a completely different user relationship, there's got to be some sort of proxy device, or transcoder for normal Internet content to go through, and even at that, based on our testing, it still doesn't translate all that well.
CED: What are your home networking plans?
Hildenbrand: We're working on a lot of connectivity options, in the summer or late summer of '01. I'm working now on a printer for one of the USB ports. I'm also working with Sony and other companies on a tablet.
CED: I love the tablet idea. It removes all the familial obnoxiousness of having someone trying to interact with the TV while someone else just wants to watch TV.
Hildenbrand: I love it, too. Recipes—you want to do that in the kitchen. That sort of stuff. It's the same logic as TV vs. PC. You don't want a lot of text—you want something pretty snappy. So I like the idea of the Web tablet for that. PCs are different—they're for work, maybe some games. So, I view our set-top as a gateway, yes, but probably not inclusive of the PC.
CED: How does the Sony set-top relate to the Sony Playstation 2?
Hildenbrand: Playstation 2 is on its own mission now. Where the two morph, I don't know. I mean, Sony clearly understands the benefits of broadband connectivity. Our box has I-link on it, as will its Playstation 2. Right now though, they're two parallel developments. Although it's more like when you're staring down a train track. The rails appear to meet at some point.
CED: Yes, but at a point always on the horizon, always out of reach.
Hildenbrand: There you go. (laughs) But that's not to say it's not in the plan. Hopefully the two meet at some switching yard, to complete the train analogy.
CED: Will you sell the boxes at The Wiz?
Hildenbrand: One of the other side effects of a full change-out with no incremental charge is that you have to figure out some level of—what's the nice way to say it? We have to do this in a very gentle way. But there's some level of non-optionalness here. All of that makes it pretty hard to say, "come in and buy one." However, The Wiz will be key as a distribution center for it, as will every other means we have. And The Wiz will certainly play a big role in add-on devices.
CED: You mean the things that hang off the USB and 1394 ports?
CED: Say I live in one of your systems, and it's next summer. I have your mambo box. I turn on the TV. What do I see?
Hildenbrand: You'll have a program listing, some kind of guide—we haven't decided on whose yet. There's a home screen. But we're not sure how the box will start up: Whether or not the subscriber can elect how the box comes up, whether or not it comes up to the last place you were, when you were watching TV, or whether it comes up on an HTML page, that's one of the purposes of the pilot tests. All the capability is there to do three or four different things, but we're not locking and loading yet. God, I feel like Regis right now. "It's all there for you." (laughs)
CED: This doesn't have to be your final answer, but, what about economics? Last year, AT&T told investors it could get something like $120 out of the box by '04. Does that math work for you?
Hildenbrand: I won't, can't and shouldn't talk numbers. But, even with back-of-the-envelope modeling, you can turn over the investment in this box, within a three- to four-year horizon, given all the new service potentials, removal of theft and the like.
CED: Including what revenues?
Hildenbrand:That's just with reduced theft, VOD, some light e-commerce, and if we offer some kind of voice or second-line service.
CED: No interactive ad revenues built in?
Hildenbrand:That's without ascribing value to interactive ads, which are hard to quantify. You know the value is potentially there, but it's hard to put your hands around it yet, because there's no critical mass.
CED: I assume with the Spyglass browser, you're looking at an ATVEF model for interactive content?
Hildenbrand:Yes. ATVEF support comes online for us early next year—not with the things we do this November/December. It requires the next revision of the Spyglass software, which will be downloaded to the boxes.
CED: Speaking of Spyglass—now that it's a subsidiary of OpenTV, does that open up any Cablevision doors for OpenTV's middleware?
Hildenbrand: No, not right away, anyway. I mean, we're talking to them. We have a good relationship with them. Sony's building the middleware for this platform now, and for us, "open" is the key.We'll keep talking. I've learned to never say never.