Taking a Look at the Thick and Thin of It
Interactive television is certainly no exception to that particular adage.
The industry is abuzz about thick and thin set-top boxes (STBs).
Nothing in life is easy. Interactive television is certainly no exception to that particular adage. The industry is trying its darnedest to focus on thick and thin set-top boxes (STBs), and what they can do to make interactivity real.
It's all about interactive capability. When it comes to interactive applications and services in the thin-client or low-end boxes (essentially Motorola's DCT-2000 series boxes), much of the heavy lifting is done by the server in the headend. With the thick-client or high-end box, very little work is done in the headend. Instead, raw code comes through the network and the box does most of the work.
While the company maintains otherwise, industry observers felt AT&T's "re-emphasis" of its commitment to develop DCT-2000 interactivity this past spring was, in fact, a major realignment in the interactive TV space. AT&T, the observers said, was shifting course and its resources away from thick-client box development in order to wring as much interactive juice (and new revenues) from the 2000 platform as it could. The thin-client platform represents roughly 16 million deployed units industry wide.
AT&T's interactive reality check forced the spotlight on other operators and their plans for thick and thin client development. Suddenly, operators across the country were weighing the pros and cons of interactivity and how to get it up and running.
As is often said, no two cable systems are identical, and apparently neither is their approach to interactivity.New kid on the block
Much like others in the industry, the folks at RCN Corp., one of the country's most aggressive and successful overbuilders, think there's room enough for both thin- and thick-client boxes in the market.
Rioboli believes the thin-client box will have a limited interactive role, at least on RCN's network. He admits he's still in the planning and evaluation stage of set-top interactivity, but at the same time, he's not all that enamored with the thin-client applications he's seen so far. "The applications aren't that impressive. And because you have limited processing and memory in those boxes, you have fairly large servers and systems that are needed to support them," he says. About the only thing the thin-client boxes are good for, says Rioboli, is video-on-demand (VOD).
At this point in time, when RCN talks about interactivity, it's talking about thick-client boxes (Motorola's DCT- 5000). But Rioboli says the challenge with the thick-client box centers around its cost, which industry estimates tag at about $450 to $500 each.
Says Rioboli, "If you look at the applications and services we're looking to put on those boxes and the revenue streams they're going to produce, and the additional capital we have to spend on the higher end boxes, the economics are a definite challenge."
What's an operator to do? Rioboli says there are two possible "solutions" to the price problem. One is to move down a notch, more or less, and use a "mid-range" box that's "not necessarily a thick, thick box that's very expensive, but instead maybe a medium box (that) will be able to do most of the functionality of the high-end box at a much more economic price." He says several STB manufacturers are already touting such a box.
The other solution looks to Europe and the high-volume boxes that have already been produced there. "In other countries," says Rioboli, "there's a standard called DVB (Digital Video Broadcast). We're looking at those boxes as well as an alternative to just going down the standard path (here in the U.S.). We think there's potential if you leverage some of those worldwide volumes. We may be able to get better, more cost-effective mid-range boxes. So, we're investigating that alternative as well."
Another advantage to the DVB approach, says Rioboli, is that it's an open system, much like the open data standard (i.e., DOCSIS). "We're the only country where it's essentially a monopoly based on which solution you choose," he says. "It's essentially a closed system. If we choose a Motorola system with their conditional access, we have to buy all of our set-top boxes from Motorola.
"In other countries, you put in a standard DVB-based conditional access system and then you go shop around to all the set-top box vendors that are interested in your business. This is much like the way the Internet technologies work. I can go buy a router from anyone, like Cisco or whatever, and I can go buy other equipment like CMTSs and modems, from other companies."
He admits that there may be some interfaces and/or components that may have to be swapped out to make the DVB boxes compatible for use in this country, but he doesn't consider that to be too much of a hurdle to overcome. "The greatest advantage," says Rioboli, "is that the set-top box vendors have to vie for your business in a (more) competitive environment. Therefore, there's going to be much more price pressure on them and they're going to be more inclined to innovate and develop new products to maintain your business.
"If you go to other countries, for the processing power, the memory, and all the capabilities of a set-top box in those countries and compare it to the same box here in the United States, they're paying significantly less there for the same functionality we're paying for over here."
Rioboli has already talked to a number of set-top box vendors about the idea, and while he can't be specific, he says there's definite interest to provide such a box. "The set-top box manufacturers that haven't been able to penetrate the U.S. set-top box market because it's been very much monopolized by Motorola and S-A, are very interested," says Rioboli.
What kind of an impact would a DVB box have in the United States? "It would be like setting off a grenade in a room," claims Arthur Orduna, vice president of marketing at Canal + Technologies. "I think it would be good for this marketplace. I think it would scare the crap out of people."
Orduna notes that there are at least 4.5 million DVB/MHP (multimedia home platform)-compliant boxes slated for production in Europe over the next few years. The question, he says, is whether such a box would work here without using either Motorola's or S-A's conditional access. He thinks it's possible if those two companies were allowed to get PowerKey or Harmony within that system.
Bottom line, Orduna agrees with Rioboli that the main attraction to a DVB box is money. "The main reason to consider it," he says, "is that you get economy of scale in set-tops.
"Here's the deal. Remember, there are 4.5 million boxes coming over the next couple of years in Western Europe. The OEMs are already lined up and delivering prototypes today. They would love to take that same DVB/MHP-compliant box, which is already subsidized (by the European orders), and turn around to the U.S. cable market and say, 'How would you like to have a box that is as powerful or even more powerful than Ultimate TV, because it also has the two-tuner capability on top of all that, and a DOCSIS return path, etc., and is significantly, by order of magnitude, cheaper than the equivalent that you could get from S-A or Motorola'?"
Orduna believes there are other incentives for having such a box in the United States. He notes a DVB/MHP box means Java is the execution environment and HTML is the presentation environment, which in turn "means you can create applications real fast." This ability to create applications almost on the fly, says Orduna, would also have a huge impact on the way operators could generate local revenue.
"Is there a simple application where they can create an interactive billboard for local advertisers?" asks Orduna. "Is there an easy template they could take out to Frank's Pizza Parlor and (use to) create an interactive commercial? It may not be as sexy as a VOD/DVD application, but I bet you can make a hell of a lot of money if you had a simple, fast way of creating iTV applications on the local level. That's another reason why DVB/MHP is very attractive."
Rioboli says RCN's middleware preference rests on the box it chooses. "We've evaluated of a lot of them," he says. "But, we're in the final stages of trying to finalize our strategy here for the set-top box, and that's going to impact the middleware (decision). The overall platform, whether we remain on the Motorola platform or go on a DVB platform, will impact the middleware decision."Thin is in...for now
Mark Hess, vice president of digital television for Comcast Cable Communications, believes the internal thick/thin "debate" that is going on at many cable headquarters around the country can be likened to another experience everyone has gone through growing up.
Says Hess: "Our mantra around here has been more or less that interactive TV is sort of a crawl-walk-run experience. You want to manage your capital spending. You want to manage the customer experience. You want to manage the rollout of these things. It fits a strategy to do something on the thin client while you're still working on the thick client set-top."
He, too, has a problem with the thick-client cost. But he also believes it's something that will eventually be addressed. "Part of the problem I see with a $450-plus set-top," says Hess, "is not only do I have a fairly large incremental cost in the home over the current digital box, but I've also got large incremental costs in the backend architectures such as servers and whatnot.
"So, if I can get that cost down so that the interactive television applications and products are really paying for the backend rather than the front end, I already know the economics of putting a digital set-top in the home are pretty good. We like it. We'd do it."
Hess says he still considers a thick-client box a moving target, what with the new chip sets that are being developed. "So, we're not looking to deploy a lot of thick clients in the near-term future."
Instead, he says, "We're going to do something with those (2000) set-tops. We're going to work on ways to make them interactive. I think a couple of things have happened that make interactivity on the DCT-2000, specifically, a lot easier. One is our agreement with TV Guide that will allow us to integrate other applications. Prior to this, TV Guide controlled the resources of the set-top."Thick man walking
If you called David Housman thick, the vice president of corporate development and new technology for Charter Communications would probably smile and thank you for the compliment.
He acknowledges his company has and will continue to deploy the DCT-2000 and Explorer 3100, but it's not where he'd really like to be. Says Housman, "We're still in the thin-client environment because that's the only thing available. We're with the rest of them (other MSOs) and it's frustrating. At the same time, we will be deploying thick products. In fact, we'll be commercially deployed by the end of this year with thick products."
Housman says the company will deploy Motorola's DCT-5000 first in the St. Louis market. He says he's been trialing the product there for the past 18 months "getting quite a bit of experience with it."
Housman admits the cost of the 5000 is anything but comfortable. Yet he thinks it's something that can be dealt with. "Oh yeah, it's painful," he says. "But, at the same time, it's about whether you can get a business model to work. The answer is 'Yes.'
"We look at the thick box as the beginning of a lot of future applications. I don't think there is anybody who can make a business model work just for Internet television.
"I think the thick box is what enables a DVR (digital video recorder) experience and a lot of those types of applications. That's why we still look at this as an entertainment experience, not an Internet experience, because that's not what we want it to be."
Housman is also a firm believer that there's plenty of room for several types of boxes and their varying applications in the network and the home.
"We see consumers having both. They're going to have a thick client in one room and probably a couple of thins in others. In the same way we have a DCT-1200, a 2000, and probably a 2500 in the same home today, the picture down the road may look like a 5400, a 5100 and a 2500 (in the home). There are all kinds of possibilities. We're going to provide a series of offerings, and through time, our customers are going to tell us what they want and what they like."One step at a time
When it comes to digital box deployments, Cox Communications certainly has nothing to be ashamed about. Chris Bowick, senior vice president of engineering and CTO at Cox, says the company has just over 1 million digital customers, with an overall penetration of about 12.5 percent. The current split, he says, is about 60/40 Motorola and S-A, with it moving toward a more equitable 50/50 split over time.
Bowick says Cox is taking a measured approach to the thick and thin situation. He says he plans to continue rolling out the thin-client type boxes from Motorola and S-A, with the intention of adding another vendor "in the not-too-distant future."
That not-too-distant future is also a place where he sees a thick-client box in customer homes.
"We're trying to take a more holistic view of interactive television and advanced set-top boxes by not putting the cart before the horse," says Bowick. "We want to make sure we understand the business and we want to make sure that any platform and any application on that platform we choose would not only be a good business for us, but be stable as well."
That doesn't mean he thinks the thin-client boxes should remain static. In fact, he's pleased with what he's seen on application development for the 2000 and he'd like to encourage third-party vendors to continue their development work.
One thing that intrigues him is PVR (personal video recorder) functionality that he'd "love to see in a box sometime next year." Bowick says Cox is eyeing a sidecar approach, but says "a separate box is yet another box in the home that makes for difficult integration."
However, there is one thing about PVR functionality he believes needs to be addressed. "Of all the electronics that you might put in a set-top, a hard drive is the least reliable," says Bowick. "It's a mechanical assembly. And so a lot of manufacturers are looking at ways to insure that their overall set-top reliability does not decrease as a result of adding that functionality. You could provide a set-top with a disk drive that's easily removed and added back, in case there is a failure of the drive itself."Show me the money
To hear Bob Hamilton, principal engineer at Shaw Cable Systems talk, you'd swear there's some good, old Scottish blood in his veins. He says Shaw has deployed about 275,000 digital boxes. Most of them, he says, are Motorola DCT- 2000s.
So far, he says, they're "not doing a whole lot" with them, essentially confining activity on the boxes to TV Guide and some pay-per-view. "We've been looking to do something with them for probably about four years now," he says.
"The technology is there to do something. It seems you might be able to do a lot of things with them. But so far, we haven't really figured anything out that's actually going to generate any revenue."
Hamilton says he's heard the talk about e-mail, chat and instant messaging functionality. While that may hold some interest for others, he doesn't believe it really applies in his situation, having just cracked the 600,000-cable modem subscriber mark.
"You realize we have a large number of cable modems out there, so (that means) I have to give e-mail free with the set-top because it comes free with the modem," says Hamilton. "If you're going to surf the Web, you're not going to do it on the TV, you're going to do that with your PC and cable modem. So that's when you start thrashing (about) and wondering what services can actually generate any money to pay for the infrastructure."
While Hamilton might not be bullish on the thin-client platform, he does hold out some hope for the thick-client approach to interactive television. He says he's been "playing" with the thick client platform for nearly 18 months, working with VOD and billing system integration.
With his enviable penchant for demanding that services actually generate revenue, he's not about to fool himself that the thick client box is, by any stretch of the imagination, cheap.
"It becomes a more expensive chunk of hardware," says Hamilton, "and it becomes a lot more difficult to figure out how you pay for this thing. I think we're leaning toward not going with the thick client box until we have a PVR, which effectively should be by the end of this year, or very early next year." That, he says, gives you the ability to provide the consumer something new.
Hamilton also believes that the thick-client boxes, which he expects to put in a test market in the first quarter of next year, may serve as a launching pad for thin-client interactivity.
"A lot of people," says Hamilton, "are jumping trying to get in (to thin client interactivity). But, if you can get it to work on a thick client, if you find the killer app, then you just have to figure out how to take it backwards into the thin client market.
"That way, you're not struggling on the thin client, which (has) restrictions as to what you can do and how you can make it work. With the thick client, you have a lot more flexibility as to what you can do there."From the bottom up?
"The biggest challenge," says Lev, "is how not to get locked into a low-end or lowest common denominator solution when you want to evolve to the advanced set-top boxes.
"In other words, when they're clicking through on the box in the bedroom, it has to have a similar look, feel, design and interactivity as (the box) in the living room, except the living room may have more ambient movement, as opposed to static pictures, for example."
Lev says application developers and the operators who will be offering the application or service may not have to think about specific features that would be on an advanced box when they're developing an low-end application. What is important, he says, is to develop applications that give a maximum amount of flexibility to work in a multi-box environment.
This thin-first/thick-second approach to application development, says David Novak, director of marketing at Pace Micro Technology Americas, not only protects the operator, it can serve as a useful way to educate consumers.
"If you look at Comcast," says Novak, "it likes to deploy services in gradual steps for a lot of different reasons. For one, you prove the business case of the applications you deploy without taking too much of a risk in deploying apps that you're not sure of completely. If those prove successful, then you can move on to more advanced set-tops.
"That also educates the consumer. Consumers won't understand a lot of different applications launched to them at one time. If you do it in steps and educate them along the way, the more advanced applications will be accepted more readily down the road."
One of the ways to develop thin-client applications, says Steve Reynolds, CTO at Intellocity Inc., a Denver-based applications developer, is to focus on what he calls "lightweight" interactivity, which includes interactive advertising, portals, information sites, or "things that don't require a lot of heavyweight user interaction."
Reynolds says these are things that can be driven by the remote control, delivered in a broadcast mode, or things that can leverage the store-and-forward networks that ops already have in place. As an example, he cites an application for ESPN that launched on the DirecTV platform–Channel #490 on its box–called ESPN Today. It's a sports information on-demand site where users can get information on sports, stats, game recaps, etc. This interactive application, he says, is not based on any heavyweight network, middleware or DOCSIS interactivity.
Even gaming, says Reynolds, is a thin-client possibility. "There are mechanisms we can use," he says, "to push the gaming content down in a broadcast mode to the set-top box and then let the user interactively navigate through the game." Question-and-answer type games can also work, he says, and the scoring is kept on the set-top box side and then uploaded and aggregated later with other scoring information.An integrated solution?
Christopher Cunningham, vice president of business development at Wink Communications, says that the integration of certain applications on a thin-client set-top box is not as simple as just porting it to a common middleware platform. "Everyone says they can fit in this or that box, but no one says with whom," says Cunningham.
"At the DCT-2000 level, there are certain middlewares that can fit if they're the only thing that's running in that box. Some people will say that they've ported to such-and-such box. That's great, but they haven't ported along side all of these other services."
So, who decides who integrates with whom–the STB manufacturer or the operator? Cunningham thinks it's both. "In some cases," says Cunningham, "operators have already given a lot of the (box) power to TV Guide. So, they've guaranteed x-amount of memory and available processing power for the TV Guide application.
"The cable operator will say he wants a certain VOD software in there, he wants a certain enhanced broadcasting software, etc., in there. And then he'll say, 'Oh, by the way, you'll all have to fit together.'
"That's what is happening. It not only requires them to integrate with each other and maintain a consistent look and feel, they also have to decide who has control of the on-screen display, etc."
Michael Collette, senior vice president of marketing at OpenTV, thinks the integration specifications should be set by the operator, not by the EPG/IPG provider or the application developer. "That's really critical because you wind up with two alternatives," says Collette. "TV Guide, of course, would like to retain a controlling position in the market. So, it'll dictate a side-by-side integration, which means its whole code set is effectively independent of the middleware set.
"What the MSO wants, because he really wants control over his business and the set-top, is what's called the stacked model. That means the TV Guide EPG/IPG resides as an application just like all other applications on top of the middleware. And then the other applications–VOD, e-mail, e-commerce application, etc., are loaded independently."
"If you think about different applications scenarios, e.g., an EPG, VOD, or e-mail, you're typically not doing all those things at the same time. If you're watching VOD and controlling it, you're probably not writing an e-mail at the same time. So, the strategy of having swap-in and swap-out works pretty well for a large class of applications that don't require simultaneous, active participation."Content ROI is the real king
Mark Gurvey, vice president of sales and marketing at Pioneer, thinks integration problems, whether they're on the thick- or thin-client platform, take a back seat to the importance of content. "The challenge in integration is a secondary challenge to the business analysis," says Johnson. "If the ROI, based on the content offered, is not there, that becomes your number-one obstacle. That relates back to quality and desirability of content. That's the key.
"There's really no doubt that content is king. After that, everything falls into place. The lack of compelling content becomes the one obstacle that kills everything else."
Canal's Orduna concurs, but has a proviso for operators and set-top box manufacturers alike. He's had marathon meetings with programmers and content developers, trying to determine what they're looking for. "All of them," says Orduna, "are making plans for the DCT-2000. All of them are trying to get their hands on a DCT-2000 developer base. For them, the 2000 is an addressable box that they want to get into their labs to play around with and spend the next couple of months figuring out what they can do with it. All of them are resigned to that fact."
However, says Orduna, they'd like something in return, "What was common through all of it, what they were saying was, 'We'll deal with what the industry tells us is the platform. As long as the damn thing doesn't change on us and has at least a base of 10 million eyes. Because I need at least a 10 million advertising base to make part of my revenue model work.' That way, their investment and the expense is justified.'"
Just what kind of content will work? Bob Van Orden, vice president of product strategy at Scientific-Atlanta, says S-A has done a good deal of research in this area. "A couple things are really clear," says Van Orden. "On-demand programming, whether it's movies or subscription on-demand, is going to be a huge hit. There are just no two ways about it."
Another category, for lack of a better term, he calls "communications." Things like e-mail, chat and instant messaging, he notes, are enormously popular. Van Orden believes that if operators can put that experience "very simply and elegantly on the TV," it will be a huge hit.
Van Orden believes there are two other interactive content categories that will score with subscribers. "There are different names for both," he says. "One of them would be called either information-on-demand, or sometimes you hear the term 'walled garden.' The basic information that most people are interested in– weather, news, sports, stock quotes–if you can do that and not make it look like a PC, we think that will be popular.
Another application, he says, is enhanced TV. "This would include enhancing existing video content with additional information and graphics. So, if you're watching a sports program, you can pull up information about the program, the team, or the players you're watching."
One other application, says Van Orden, while not formally studied, has had considerable success wherever it's been offered. "It's kind of a wild card," says Van Orden, "but we've seen some pretty good results with it. It's what some may call food-on-demand. That's where you can order food from a local restaurant through an interactive TV application. While there's not much consensus on it, the places we've seen it and tried it, it's done pretty well."Making a killing
Bottom line, says Sid Gregory, senior vice president of engineering at Intellocity Inc., is that interactivity has to address what the operator feels is most pressing. Anything else could be a big waste of time, talent and money.
"Everybody always asks, what's the killer application," says Gregory. "How often have we heard that the last five years or so? But I'm not sure there is a killer application.
"I think the counsel that makes most sense for operators is to say, let's make sure to help them focus on the real business problem they're trying to solve. Is it churn? Is it ingress or competition by an overbuilder or satellite operator? And, what's the most economical investment he can do to address that problem?
"Is it in fact as simple as stepping up and introducing VOD? Is it more complex than that, where the overbuilder is promising everything under the sun? It comes down to just how competitive operators want to be, and at what price point are they willing to play."