I want to put a bit of meat on the bones of my previous post about ads and content by sketching out a hypothetical future YouTube product that could make Google a big pile of money. This also relates to the discussion Jerry and I had about whether the Internet will kill broadcast TV.
People in the future are not going to want to watch TV at their desks. Couches will be as comfortable as they are today. So imagine Google sells a $99 “YouTube TV” set-top box that plugs into an ethernet jack and streams video from the Internet. It would probably look something like Apple’s forthcoming iTV set-top box.
You would have a YouTube account, and you’d have a variety of ways to select content that you want to watch. you could subscribe to a television show or video podcast (which will likely be one and the same), so that each week’s episode automatically appeared in your queue. As you were browsing the web at work, there’d be a little button at the bottom of every YouTube video (including embedded ones) allowing you to place a video in her queue for later viewing. The system might use collaborative filtering to suggest new programs you to watch, as TiVo does today. You might also give certain friends permission to add videos to your queue.
When you sat down on the couch and turned on your YouTube box, it would automatically start playing the first program in your queue. The remote would have three buttons that were green, yellow, and red. Pushing the yellow button would signal to the machine “this looks interesting but I don’t want to watch it now. That would put it back in the queue for viewing at a later time. The red button would say “This sucks. Never show it to me again.”
The system would automatically insert commercials in between segments of the programs. But they’d be different from traditional commercials in several important ways. First, they would be better targeted. An advertiser could pay to have his ads in front of, say, 30-somethings who watch The Daily Show and The Sopranos. Because they’re more targetted, there would be fewer of them. A company selling denture cleaners will pay almost as much money to get his ad in front of the 30 percent of viewers over 65 as he will to get it in front of all viewers.
The system would also keep track of which ads you had seen and not subject you to the same ad over and over again. For example, Apple could produce a series, of ads, and ensure that each of them gets shown to each viewer exactly once. Or an advertiser could specify that each customer should only see a given commercial no more than once a week. Currently, advertisers over-buy ads because that’s the only way they can hit every single consumer. Obviously, better targeting would benefit the company, the viewer, and YouTube.
And the red and yellow buttons would work exactly the same with ads as with content. Pushing the red button would tell the system to never show you that ad again. Pushing the yellow button would skip the ad, but the system might show you the ad again in the future. And pushing the green button would trigger an advertiser-determined action: it might show you a longer commercial with more information about the product advertised, it might send you or a friend an email with details about the product, or it might bring up some kind of interactive content. Or, if the ads are part of a brand-building series, such as Apple’s Mac vs. PC spots, it might show the user the next ad in the series.
Not only would this minimize the perception that ads were intrusive and annoying, but data about which ads got red- and green- buttoned by which consumers would be a gold mine for both YouTube and their advertisers. For example, if a customer red-buttons 5 car commercials in a row, the system might not show that customer another car commercial that month. Conversely, a consumer who clicked the green button on more than one car commercial could be flagged as interested in cars, and more car commercials would be added to the rotation.
Likewise, advertisers trying to make entertaining brand-building commercials, such as Apple or MasterCard (with its “priceless” campaign) would find it extremely useful to know which of their commercials gets red- and green- buttoned most frequently. The least popular ads can be retired, and the most popular ones put on heavy rotation. Indeed, they can afford to take a lot more risks with their commercials. Instead of making five and playing them equally, they can make fifty, play each of them with a small sample of viewers, and then expand the ones that get the best feedback. They could even have contests where they invite the amateur filmmakers of the world to submit commercials, give each of them a small test run, and award cash prizes to the best-performing commercials in exchange for the right to run them more widely.
YouTube would also likely charge advertisers by the second, giving them an incentive to make the shortest possible commercials. This would encourage advertisers to produce short “teaser” ads: “Ford has just unveiled the 2016 F-150 pickups. Click the green button to learn about its new self-driving feature.” They could buy three ten-second teasers for the same price as a full 30-second ad. A commercial break might consist of 3 ten-second ads inviting the user to click green to learn more about various products. If the user wasn’t interested in any of them, the program would resume after 30 seconds.
The indiscriminate, one-to-many nature of television is a limitation that’s bad for viewers and advertisers. Advertisers don’t want to pester people outside of their target audience with ads. And on the flip side, most people actually do like buying things–it’s just that most commercials aren’t advertising anything they’re in the market for. The traditional view of advertising as a nuisance we put up with to get the content we want simply reflects the limitations of 20th century broadcast technologies. It would be perfectly feasible to write software to strip out Google’s contextual ads, but I’ve never heard of anyone doing so, because if anything they enhance the value of Google’s search engine. I see no reason you couldn’t design a video ad system that’s similarly non-annoying to users.