As the inventor of switched digital broadcast technology, BigBand Networks made a name for itself among the country’s biggest cable operators, crossing over to win Verizon as a customer before it went public earlier this year. Telephony’s Ed Gubbins recently spoke with the equipment vendor’s chief IPTV architect, John Reister, who is tasked with further penetrating the telco market. His vision for the future of the IPTV business is as big and clear as a high-definition video on a plasma screen.

On the future of TV: TV is going to look a lot different. With different camera angles when you watch NASCAR and baseball games, you’ll be able to bring up views of the pitcher and the catcher and the outfield, all at the same time. The program guides are going to make recommendations to you along the lines of what Netflix and Amazon.com do: “You watched these 20 shows over the last several weeks, here’s what people with similar interests are watching at this moment.” We’ll see even more channels in the future, but rather than the perception that there’s nothing on, it’s going to come with search features that will highlight the things you’re interested in. You’ll have more interactivity. You’ll have advertising specifically targeted to you. Addressable advertising is going to be massive.

On the future of TV advertising: Today, advertisers place ads through agencies based on program ratings by demographic category. If they’re trying to place an ad that appeals to families with children, the agency will tell them, “You want this channel, this particular show.” All that is done predictively. You buy based on Nielsen ratings, you pay a cost per thousand views. That will be completely turned on its head. In the very near term, an advertiser can place a national ad but say, “I only want to hit the demographic of shotguns and pickups for this ad, and I want to go for the upper crust with this other ad.” They could choose five ads to place in the same slot. The service provider can sell the same slot over and over again because it can go to different demographic categories.

You can also place ads by ZIP code and cross the two together. This targeting is done in real time so you know—and pay only for—the exact number of viewers who are watching. Completion data today is pathetic. If an advertiser wants feedback on an ad’s effectiveness, they have to conduct polls. With this new model, we can report in a minute how many set-top boxes in the demographic and ZIP code saw the ad. We can indicate if someone changed the channel. You’d get immediate feedback so you can make a substitution in real time.

On time-shifted TV ads: I think we’ll see time-shift TV and network [personal video recording (PVR)] balanced against home PVR. Advertisers can’t stand home PVR because they create ads to be viewed at a certain time. McDonald’s wants to do ads around dinnertime. If Macy’s is having a white sale on Saturday, and you record it and play it back on Sunday, the ad’s completely useless. One of the benefits of time-shift TV and network PVR is you can put the ad in at the time it’s being played, so you’re certain it’s going to be relevant. Ironically, advertisers have been fighting [network PVR] aggressively. But advertisers would really like network PVR—putting ads in at the time of playback. Content owners don’t want it to happen because they’re worried they’ll lose control of the advertising.

Home PVR disrupts [the traditional advertising model], but content providers shrug and say that’s out of their control. Video providers want to record the programs on servers inside their networks so subscribers can access them with time-shift TV or network PVR, but content providers have sued to prevent that from happening because advertisers told them [the ads] had to be shown at a certain time. It’s ironic to me: If you don’t allow network PVR to happen, you’ll end up forcing everyone toward home PVR, the direction they’re already going.

The last piece of this complicated puzzle is: If the ads are placed in at the time of playback, the company placing the ads would be the one doing the playback: the service provider. [Content owners] want to sell the ads themselves. So you’ve got tension over who sells the ad. This whole thing has to be worked out so you can get the benefit of relevancy.

On interactivity and TV-commerce: You can put clickable parts in personalized ads targeted to specific individuals. A little note will come up saying, “Click on this.” If they click on it, the sky’s the limit in terms of what you can do: go to a URL, launch into video-on-demand, make a purchase, enter a contest. You could get e-mailed a coupon for 10% off your next visit. PCCW in Hong Kong has an electronic program guide that’s a little like iTunes. It lets you bring up movie trailers. If you want to see the movie, you click on it; it will bring up listings for local theaters. You can buy a ticket. It’s like Web-based shopping, but you’re doing it from your TV using a remote control.

On ensuring IPTV quality: When you deliver high-definition IPTV, you’ve got sustained, high-bandwidth flow that bounces around quite a bit in terms of the bandwidth it consumes. It’s quite hard to deliver something high quality. When you’ve got upgraded wireline networks, it helps. But we’ve got WiMAX coming along and mobile video to handsets. The bandwidth available to the user may fluctuate, so it’s a massively difficult problem.

A big part of that problem is when you watch video on the Web, there’s a delay. If the buffer stops filling because of a network issue, your picture freezes until more video comes down. That’s completely unacceptable on TV. You have to deliver video in real time. Traditionally, in telco packet networks, packets go into a queue, and when there’s not enough bandwidth, they back up, and eventually the queue overflows, and you drop packets. With video, that’s unacceptable. You can’t send only 25 seconds of video and have it spread out over 30 seconds; 25 seconds of video have to be delivered over 25 seconds. The solution is more application-layer intelligence at the edge of the network—being able to examine the video and adjust the rate dynamically to match the available bandwidth.