Splicing together cable's advertising future
As vice president of technology for Comcast Spotlight, the advertising sales operation of the nation's largest MSO, Paul Woidke knows a thing (or 10) about the engine rooms of the advertising business. And it's a technology that matters to the business: Ad sales will post some $1.5 billion in unregulated revenues to Comcast's balance sheet this year. That's about 6 percent of the MSO's total revenues—but about 20 percent in free cash flow, the number scrutinized by Wall Street observers. In this edited Q&A with
CED technology analyst Leslie Ellis, Woidke discusses how to get ready to insert ads into digital simulcasts, what on-demand is doing to the advertising business, and how the fun is just beginning when it comes to cable and local ad sales.
CED: What are the big topics in your world right now?
Woidke: Probably the best thing to happen over the last three years is the wide dispersion of interconnections among operators. If you go back four or five years, there were maybe 10 to 12 interconnected markets in the country. Now, virtually all markets are interconnected. The markets [are] operated by Comcast, Time Warner, and a large number of other operators.
CED: What does that do for you?
Woidke: An interconnected market lets advertisers buy cable on a functionally equivalent basis with broadcast. When you buy broadcast, you buy one station, which serves the entire market. If you look at any of the markets we operate in, buying the entire market could constitute 20, 30, 40 different ad zones. In the old days, you had to account for all of those zones separately. An interconnect means an advertiser can buy the entire market, through a single point, and not have to deal with multiple tapes, invoices, contracts or negotiations.
CED: Which opens the door for segmentation?
Woidke: Right. It's the reverse of the interconnect. People sometimes stare at me sideways when I say that: If you want to segment it, why did you interconnect it in the first place?
Say a car company wants to buy a spot to run on "Larry King Live." Until you can assure that the spot will run consistently across the interconnect, you can't disassemble the market into segments.
CED: And now? What are the most interesting things on your radar now?
Woidke: Now, it's addressable advertising. There may be 40 ad zones in Chicago. But those 40 zones reflect how the plant was laid out in the 1970s. It's not necessarily rational, at least from an advertising perspective. As we move forward into addressable advertising, it allows us to re-segment that plant and deliver a more appropriately segmented audience for an advertiser's product. The way we'll be able to do that, at Comcast, is within the push to digital simulcast.
CED: You're talking about your DPI (digital program insertion) launch?
Woidke: Yes. Digital simulcast is going to be a real alteration for the ad sales business. It means we'll be able to reach all homes within a market, with all of our ad-supported networks, delivered on a digital basis.
CED: Does DPI launch after you're up with simulcast, or during?
Woidke: Simultaneously. That's significant, and by design. It means that in any given zone, within any given DMA, all networks are offered in analog and in digital.
Advertisers don't buy analog eyeballs or digital eyeballs. They buy everyone who watches MTV, or CNN, or ESPN or Discovery.
We spent many months on this project, coordinating with John Donahue's (Comcast's SVP of engineering operations) group on the cable system side, to assure we could go forward hand-in-hand. So that as a simulcast turns on, the digital feed of MTV will contain the same advertisement that's contained in the analog feed of that network. We can verify it on a single point of insertion basis.
CED: What's a "single point of insertion basis"?
Woidke: One could imagine having insertion equipment, sitting there, putting ads into the analog networks. You leave that running over in the corner. Next, you take the digital version of all of those networks, and insert spots into them, using the SCTE 30 and 35 standards.
If we did it that way, it would mean that in each zone, we'd have to track two sets of ad insertion gear. Two sets of schedules. Then, we'd have to bring all that data back together, and parse it to marry up with each analog and digital occurrence, in order to generate an invoice. Having a single point of insertion fixes all of that.
CED: Where do you insert the ad, at the headend or the edge?
Woidke: In the design we're executing, the commercials will be encoded and stored on a single server. That server will insert the spot with DPI, into the digitally-delivered networks. After the commercial is inserted, that digital stream is delivered directly to the digital set-tops in the market. That same stream, with the ad in it, gets decoded back to analog, then distributed through that same zone in analog, to go to the analog devices.
CED: How did you architect the storage? Centralized?
Woidke: Right. Instead of having an ad server sitting in each of those 30 or 40 headends—each one a computer that someone has to take care of—I can centralize servers in fewer places than the number of headend zones I have. I can pull them into groupings that are more easily managed. Each server can be enlarged to handle four, five or six ad zones.
In a centralized architecture, you can put the splicers in the rack, next to the ad server, and do the splicing in that one facility—then send it out to the number of zones served. Or, they could put the splicers out in the zones, out at the edge.
CED: Which one are you doing?
Woidke: There are pros and cons of each, all purely architectural. It relates to the bandwidth and the cable system design for a market. From an ad sales standpoint, I'm pretty agnostic. Whether I insert centrally into five zones, or whether I insert at the edge of those five zones, I'm still able to provide the same degree of quality to the advertiser. So that's more a system-side decision, depending on bandwidth availability and plant configuration.
CED: You're first on this. What should CED's readership know, as they start mapping out their plans? What matters most?
Woidke: The absolute key thing is for the technology, engineering and operations people on the ad sales side, to be in lockstep with the technology, engineering and operations people on the cable side. There tend to be operational silos. You have the systems people on one side, and the ad sales people on the other. It's not that they dislike each other, but they just tend not to directly communicate.
When it comes to digital simulcast, to do it correctly and smoothly, with no bumps in the road, it's critical that those groups be talking. The moment the planning begins, a dialogue [should occur]: How are we going to do this; how do we keep moving forward together? It's been critical to how we've gone about designing simulcast, and beginning the implementation.
CED: Does everything exist that you need, technologically?
Woidke: There are things that need refinement and enhancement. This whole thing has been a rapid call to arms, and the vendor community rose to the occasion.
CED: What's your position on DVR as an advertising killer?
Woidke: It isn't just the DVR. It's VOD, too. Of course, there have been plenty of people telling us that on-demand will kill advertising. We'd be foolish as an industry to not recognize that it'll have an impact.
Let's also recognize that DVRs have not reached predicted penetrations. Four years ago, reliable and well thought of research companies were projecting that by 2005, there would be close to 40 million DVR devices installed in people's homes. It's not even 20 percent of that now.
Cable television has been providing choice for the last 10, 12 years, with an ever-escalating number of channels for the picking. The issue of control is the next one that comes along. Control means, it's nice that I have the choice of those channels, but can I watch them when I want?
CED: Nonetheless, those who do have DVRs tend to skip commercials fairly regularly.
Woidke: Yes, and at a very high rate—70 to 90 percent of the time. I'm not...sticking my head in the sand and saying, "this too shall pass." It has to be addressed. But there is a flip side to this: People are fast-forwarding through spot advertising for two reasons.
One is, you can't look at what the industry has done, taking six to eight minutes every half hour, in two-minute pods, to stuff ads in the middle of programming, and not expect viewers to do something about it. Look at TV in the 1950s and 1960s. There wasn't that volume of advertising and clutter in programming. If we haven't killed the goose that laid the golden egg, we've at least been chasing it around with a two-by-four.
CED: And the second reason people skip ads?
Woidke: If you're repeatedly showing advertising for products and services people don't care about, need, or want—it's no different than reading a magazine or a newspaper. You turn the page. When you go to the mailbox, you're not going to look at the catalog for the parakeet, if you own a dog. You throw it away. People are skipping over the ads because they're irrelevant to them.
But when an ad is relevant to someone, it isn't an ad. It's information. People strive and yearn to have information. That suggests that with DVRs and on-demand, there is the opportunity to provide people with information about products and services that are relevant, that are significant to their lives. In doing so, we provide a service both to the ad community, because they reach the people who care, and to viewers, because they get desired information.
CED: What are your thoughts around the notion of "bound applications," where a trigger comes along with a TV show—or a commercial?
Woidke: We're certainly supportive. We view it as an opportunity to create linkages for advertisers and for programmers. Advantages that accrue to the programming community that could well enhance their relationship with their advertisers.
CED: How so?
Woidke: I'll give you an example. This weekend, I was watching home improvement programs, like I often do. I thought, wouldn't it be great if, at the end of the show, I could use an enhanced linking mechanism to find another episode, or something like it, by that same programmer—maybe stored on a VOD server? That network could then create a brand image of their program. They've strengthened that brand, by letting people go watch it in an on-demand environment. It's quantity with control.
CED: How does your work on segmentation stack up against satellite competitors? Can the spot beam keep up, from an advertising perspective?
Woidke: If you were to take the 210 DMAs that exist in the country and say there are 40 ad-supported networks (to pull a number out of my ear), then say, within each of those, I want to target within 10 different zones, right there, you're beyond what satellite can do. They are the cannon; we are the .22. There's just no way they can deliver the degree of segmentation that we can, without having to invest a significant amount of additional capital.
We've built it out. We're done with that. Now, it's implementing on top of the infrastructure that's already there. We're going to be able to do this targeting, in the 40 or 60 ad-supported networks, including analog, SD (standard definition) and HD (high definition).
CED: Switched broadcasts and digital advertising—how do you see them intersecting over time?
Woidke: We continue to look at switched broadcast, to understand what opportunities might be there. The thing to remember about switched is, if nobody in your node is watching HGTV, and if you turn it on, that stream gets turned on and delivered to you. But if I live five blocks away, and I tune in, I join your stream. The first person to join has a one-on-one on-demand experience, but the second person joins the first person. I'd have to think about how to segment differently between me and you, watching HDTV on a Saturday afternoon.
CED: Any closing thoughts to this readership?
Woidke: If I were to say something to the technology heavy-lifters on the system side of the company, it's this: Ad sales isn't something that's going to go away, because of DVRs or anything else. Instead, ad sales will become far more precise and relevant.
Advertisers are not going to stop trying to reach the public to talk about their products [just] because it's more difficult now. They're not going to say, forget it, we're only doing newspapers. Or, we're only doing radio. The advertisers and the agencies that want to tell the public about their products are looking for new, innovative, creative ways to reach the public. Cable is positioned better than terrestrial and satellite TV to deliver what advertisers and TV viewers want. That's our opportunity.