Internet TV has Arrived: Coming Soon

Last summer I decided to replace my TiVo with a DirecTV DVR receiver instead. I loved my TiVo but the infrared sensor cable that stuck out from the face of the receiver to change DirecTV's channels often failed if it got bumped. So when I went to watch Monk or Lost it would turn out to be a show on another channel. Boy did that burn my biscuits. Sadly, right as I was making this decision, DirecTV ended its relationship with TiVo and began using its own DVR interface. I've been suffering ever since (I'll be OK). In the process of trying to learn DirecTV's obtuse interface I failed to record a few episodes of my favorite shows. While these instances were frustrating, it did provide me with an occasion to glimpse the future of TV. Having missed an episode of Survivor, I just hopped over to my computer and watched it for free on CBS.com. Phew, I wouldn't miss the tribal merge or who got voted off after all.

Television networks and several new online services like Joost and Veoh are starting to deliver television via the Internet. For example, Veoh contains episodes from NBC, CBS, and FOX, in addition to one hundred twenty-one other channels (and counting). ABC provides access to all its most popular shows directly through their own website using their custom player (they even have streaming HD).

It makes me happy to know that I no longer have to worry about whether or not my DVR recorded my favorite shows. But Internet TV is becoming much more than a backup plan when I miss an episode of 24. The simple act of shifting the delivery of television to the Internet radically increases the content choices and the way advertising is delivered. Both of these changes will benefit consumers. Enlarging the menu from shows airing this week to everything that's ever been aired is a nice improvement. But in this newsletter I'll focus on how Internet-based TV will improve television advertising.

As the title of this newsletter implies, there is an already/not yet aspect to today's Internet TV. We can already watch all the new shows on the Internet (we can even start watching Internet TV on our set top boxes using a few cables and adapters). But as far as the improvements to advertising, we are just on the verge of seeing these new possibilities emerge. In fact my thesis is that the current limitation of the amount of previously produced television content--all episodes of Lost and The Land of the Lost (remember Sleestaks ?)--relates to the changes emerging in Internet video advertising. I believe the current limited television available on the Internet directly corresponds to the limited implementation of these new advertising tools and techniques. I suggest that as these new advertising methods are deployed, additional television content will follow--quickly.

As it stands, if you watch a show like NBC's Last Comic Standing on Veoh, you'll have to watch pre-roll, mid-roll, and end-roll ads. In this particular case, all of these ads are the same ad. Now I don't mind watching ads in exchange for enjoying a television program. It's part of the economics of TV, and always will be. Besides, I've done it all my life. The odd thing about watching Last Comic Standing on Veoh is that every ad is AT&T announcement of it's merger with Cingular. Not five or six different ads, and not six different versions of the ad, but the very same 15-second ad every time. Or, if I watch Lost in high def on ABC's streaming Internet TV player, the entire episode is sponsored by one brand--in this case, Clorox.

When television is delivered via the air, cable, or satellite there are about 18 minutes of available advertising space to fill during a one-hour program. Advertisers line up and fill these slots. They base their ad-buying decisions on very rough estimates of total number of views and extremely generalized demographics of the audience. While I really appreciate the fact that Clorox sponsored ABC online so I can watch it for free (thanks Clorox!), my participation of viewing the ad was negligible since my wife does most of our shopping and makes the decisions about which brand of bleach to buy. (Of course here I am mentioning Clorox in this newsletter--so I guess it wasn't entirely useless in my case). But the act of broadcasting a marketing message to an entire audience in hopes of influencing a few is not necessary when content is delivered via the Internet. Internet-based television offers an environment where ads aren't broadly targeted based on a generalization of the audience, but are selectively delivered based on video content and the preferences of the individual.

Before I launch into how some of these new technologies are making this possible, it's important to see how potentially big this shift is. I've written recently about online advertising and how Google's AdWords and AdSense programs have demonstrated the amazing potential for turning the "long tail" of web content into massive and effective opportunities for advertising. Before Google, it was still pretty expensive to advertise online, and the industry was dominated by big brands placing banners on popular sites. Smaller publishers and advertisers were marginalized. But because Google was able to automate the placing of ads next to relevant search results and on related AdSense partner sites whose content was also relevant, they matched up the long tail of publishers (every blog or website that wants to participate) with the long tail of advertisers (people like me that just want to drive a little more targeted traffic to their sites). Gold.

Now, think about the current waste in television advertising. Every time an irrelevant ad is shown to an uninterested viewer, there is a wasted opportunity. But what if we weren't all seeing the same ads when we watch the same television show? What if I were ads that appealed to me as an individual--not just a resident in a certain region--during the same commercial break? This alone would drastically increase the number of advertising opportunities since wasted opportunities would be replaced with more appropriate ones. Multiply that by the amount of content we'll be able to choose from and suddenly you have a greatly expanded, more effective advertising marketplace in which almost anyone can participate.

In this light, Google's billion-dollar purchase of YouTube starts to make more sense.

Digitalsmiths is a company that indexes video content. They started out working for television and movie production studios that had large video libraries. They wanted to create searchable video databases to find clips for use in other media productions. Digitalsmiths built automated software that scans video in real time; recognizes famous actors, objects, brands, sounds, music, and dialogue; and maps these elements to relevant keywords. They used these keywords in proprietary search engines for the studios. But Digitalsmiths realized that their video indexing software could also be used to pull contextual ads from ad networks in the same way that Google includes relevant ads in search results or in their AdSense publisher network. You can think of Digitalsmith's software as the AdSense of video content.

Digitalsmith's innovation is bringing automated distribution of advertising into the online video marketplace in a relevant and contextual way. Let's look down the road just a little and imagine how this kind of technology could enable ABC, for example, to put more of its content online.

ABC is experimenting with delivering television via the Internet. But the way they're selling advertising online, the old way, doesn't take advantage of the Internet's possibilities. They get brands like Clorox, S.C. Johnson, and Principal Financial to back their online "broadcasts." I have no idea what these advertisers pay, but I do know that for this relationship to exist, a brand marketer and a network representative had to exchange a bunch of phone calls and emails. They probably made presentations to their bosses, signed contracts, and had many meetings to put the whole deal together. Then they had to coordinate the delivery of video assets, visuals, and voiceovers--hardly an automated process.

But suppose none of that had to happen. What if ABC instead ran the show through Digitalsmith's software? Rather than having Clorox sponsor the whole thing, they could display ads, not based on pre-negotiated deals, but automatically, based on the content of the show--maybe a travel package for a Hawaiian vacation since Lost is shot in Hawaii. Perhaps, as a result of a product placement in one of the character back-stories, the same product's ad could run during a mid-roll ad segment.

The ability to automatically display video ads based on a video's content would be a boon to networks like ABC because right now they have to do a lot of work to line up sponsors for their online features. If they were to put all of their content online with the old way of finding sponsorship, whose job would it be to find ad sponsors for every episode of The Jeffersons and all of their other classic shows? But if automated contextual ads could run against any video content, why shouldn't ABC benefit from all of their properties? While there may not be as much demand for The Jeffersons as there is for Lost, there is some demand. Just as almost all of Netflix 70,000 titles are in distribution to one degree or another at all times, someone somewhere is longing to watch The Jeffersons right now. In fact, just one upload of The Jeffersons title clip has been viewed 15,535 times on YouTube since April 16, 2007. Unfortunately for ABC, instead of people watching it from them so that they could reap the advertising revenue (it's their property after all), people are illegally uploading and watching it on YouTube. But imagine every episode of The Jeffersons and every other show ABC ever produced were accessible online. Multiply that by all of the other networks' shows--there's a lot of content and a lot of advertising opportunities that are being missed.

But what about Clorox's perspective? How would they benefit? If ads are automatically displayed based on video content indexing, how can a brand get maximum exposure? One answer is that any brand that wants to spend money to reach a mass audience will always have opportunities to do so. It may take new forms--product placement is one being used now--but if you have the bucks, the opportunities will be available. But while mass advertising is easy right now, no marketer wants to pay for more advertising than they need to increase sales of their product. If it takes a million views to sell a thousand units, that's what they'll buy. But if they can sell a thousand units, and only pay for ten thousand views--all the better. And the Internet's ability to connect content with relevant ads definitely maximizes an advertiser's spending.

There's one other problem from the brand's perspective. If ads are automatically indexed based on video content there exists the possibility that their brand is displayed against inappropriate content... like when Verizon's video ad was the pre-roll to the New York Times video of the hanging of Saddam Hussein. Verizon was not too pleased about that. But in the same way that technical indexing can identify content matches, it can also filter out inappropriate ad placements. ScanScout, another company bringing video content indexing software to market, has thought this through and includes a brand protect feature in their video indexing solution.

But to really maximize advertising relevancy and efficiency, there's another capability the Internet offers--and it has to do with you.

Ben Weinberger, CEO of Digitalsmiths, gave me this example... suppose a year or two from now you're watching an old episode of Seinfeld online. In this episode Kramer buys a new shirt. The video indexing engine decides that a clothing store ad would be most appropriate. It checks its clothing keywords against the ad networks. The ad network responds to the query with ads for blouses at Old Navy and suits at Saks Fifth Avenue. Which should display? Well, that's where I come in. Am I a man or a woman? Do I tend to wear T-Shirts to work or expensive suits? What is my general income level? If the site that's delivering this ad has a way to know my demographics and my specific preferences it can make a better decision about which ad to display.

But how does it get this information? Aside from my IP address, which might give them a rough idea about where I live, how does it know anything about me personally? Well, I can tell them. What?!? "You tell them?" you're thinking, aghast. That's right, I tell them.

We're all getting used to the fact that Google, credit card companies, TiVo, and all manner of websites track our visits. They plant cookies, store profiles, and recommend products based on past purchases. We're being profiled. At a certain level we have to just get used to it. But on another level we should be able to control it. After all, it's my profile, my preferences, my clickstream, my purchase history, my bookmarks, and my cookies. My digital profile is mine. Shouldn't I have the right to control it? Shouldn't I be able to add or delete elements of my profile? Shouldn't I have the choice of when, and if, I offer these details to the services I use? I say yes--and it's in everyone's best interest if we can control and use our profile intentionally.

This is the APML Workgroup argues. They're trying to develop standards for building personal profiles based on elements such as our browser history, email, documents, RSS subscriptions, and many other aspects that define our digital fingerprints so individuals can control them and websites can use them more effectively. Most of us already have various profiles out there. If you ever bought something at Amazon you have a digital profile. If you use Google desktop or iGoogle you have a digital profile. If you visit websites that plant cookies (and almost all of them do) you have a digital profile. But you can't control most of these profiles or how they are used.

But suppose you have an APML file that kept track of all these elements. Suppose you could opt to include or exclude any aspect of your profile. With your profile in hand you go to ABC.com, Veoh, Joost, YouTube or whoever to watch a video or TV show. Now, before you start watching stuff, you have the opportunity to use your APML on this site. If you do, the site will identify content that you would be most likely to be interested in, and will filter out any ads that you would certainly not be interested in. It might present you with special offers from advertisers whose products or services closely match your profile. Perhaps, since the content provider is able to more efficiently and effectively deliver the most appropriate advertising, it offers to "share the savings" in efficiency by minimizing the amount of advertising you see. Would you choose to give them access to your APML? If the site was authenticated by an organization like Attention Trust for compliance with usage and privacy standards (they don't store the info beyond your approved use; they don't sell it, etc.), would you feel more comfortable using your APML then? I would. Especially if I received other benefits and incentives as a result.

For website owners like Amazon, Google, Joost, Veoh, and YouTube it be easier and more useful to use these profiles rather than building their own limited profiles. It make their job easier and ultimately make their sites even better at presenting the most useful content and best marketing offers. And for the advertisers, this would greatly improve their targeting. And while we all love to hate ads, in reality it's the irrelevant, un-targeted, junk mail, intrusive ads that we hate. Give me a great incentive on a new car, when I'm in the market for a new car, and I don't hate that ad at all. It's a win, win, win.

I want my Internet TV. But I'll have to enjoy what's there now and wait for the rest to show up over the next couple years. In addition to the technological adoption time, there are also legal/contractual barriers to Internet TV. Today's intellectual property laws and existing long-term production contracts don't work well with the Internet as a distribution system. Contracts with content creators and actors hinder the broad distribution of existing productions. For example, Chris Anderson points out in The Long Tail that the very popular sitcom WKRP in Cincinnati was unable to go to DVD release until just recently because of how many content participants were involved in each episode. Producers, writers, actors, and directors get a cut of revenue from a production, including when it's in syndication. In the case of WKRP, there were many popular songs playing in the background of each episode--often the music was tied into the premise. Each song had to be cleared for use. The artists, producers, and labels that owned each song had to get a piece of the pie, too. In order to go to DVD, the studio had to re-negotiate new contracts for every song in every episode. In fact, as it turns out, in order to go to DVD they remastered the show to remove many of the old songs and replace them with generic music.

If going to DVD was so difficult, how much more difficult will it be for shows like WKRP to go online?

Additionally, there is a big brouhaha developing between Hollywood studios and the Writers' Guild over new proposals to standard contracts. One issue in contention is what the revenue sharing agreements will be when content goes online. Of course there's also the big fight between Viacom and YouTube over pirated video (like The Jeffersons clip) hosted on YouTube.

This is a serious matter. I believe in the rights of content owners. If they can't be protected, then advertising revenue is threatened. No ad revenue means no budgets for writers and actors and no new TV for the Internet. And as entertaining as watching a border collie rip apart a giant pumpkin on YouTube may be, I, for one, am not ready to trade excellent shows like 24, Lost, and Heroes for it (border collie lovers may disagree with me--pumpkin lovers though, are in my corner).

Right now, today, you can sign up for Joost, Veoh, or go to any of the networks' websites and watch their most popular shows on your computer. If you're a gadgety person, you can even hook up your laptop to your television and and watch Internet TV on your regular TV--Veoh even works with a remote control. Steve O'Hear has an excellent post that lists eight of the top Internet TV sites and a review of each of their services on the Last100 blog. It's called 8 Internet TV apps in 8 weeks.

I'm happy to have an alternative to my DVR for watching TV. But for now you'll only find the most recent episodes of the most popular shows, and you'll have to sit through some very untargeted ads. But soon, as video scanning technology, brand protection features, personal profile matching technology, and contractual arrangements mature, we can look forward to a growing list of online video and television content to enjoy. And I'll finally be able to bring my DirecTV dish to the dump for good.