Most internet users know by now that the ads they see online are tailored to them. The proliferation of trackers and beacons on every website, the infinitude of ”cookies” placed on your computer, the push to identify yourself every time you use any online service—these are all the result of web giants trying to optimize their advertising and make it more “relevant” to your interests.
Is it creepy? You bet. But it also makes sense. The global nature of the web makes personalization essential for advertising to work as a business model. There is little point showing someone in London an ad for a bakery in New York. A twentysomething is as likely to be interested in diapers as a fortysomething is in club nights. It’s getting more sophisticated every day, though whether it really works is up for debate.
Now advertisers want to bring that same personalization to video. And they’re going to use everything they know about you to do it, with a million different versions of the same ad ready to go depending on who’s viewing.
An ever smaller target
The first step is to identify smaller audiences. “You can start with a broad campaign reaching a lot of the US population, like a Super Bowl campaign, and then you could see smaller audiences, and smaller audiences, more personalised messages, and refining even more,” Facebook’s product dirctor, Fidji Simo, tells the Financial Times (paywall). The idea is that advertisers will be able to serve different videos to people based on whether they have already seen a previous one.
Fidji’s comments are less about a new product launch than the ongoing evolution of online media. But it is no secret that Facebook is actively pursuing video ads—or that personalization is at the core of Facebook’s advertising offerings.
The workings of such a system are fairly straightforward. Just as Facebook knows which post you liked, it knows what videos you watched. Technologies to track these interactions online already exists.
Indeed, such personalization should be a welcome move.
“People hate untargeted advertising,” Stefan Weitz, until recently director of search at Bing, told me in a broader conversation about personalized advertising last week.
“I was watching a program on Hulu and they kept running this one ad for Red Lobster [an American restaurant chain]. I have nothing against Red Lobster but I just don’t ever eat there. And the fact they ran it six times in that show, not only is it annoying, but it’s poorly targeted.”
A million ads, all same but different
Better targeting will come from better data, something big internet firms have in spades. For a sense of how that works, look to Sky, a British satellite broadcaster. Using data from a variety of sources, including national statistics, university studies, and polling firms, Sky can already target TV ads based on a long list of household attributes: How many cars do you own? Is your cellphone on contract or pay as you go? Do you read broadsheet newspapers, tabloids, or none at all? How old are your kids? Is the man in the house an ”any shirt will do” sort or a “brand boy”? Does the woman prefer “catalog classics” or does she see “image as identity”?
Accurate targeting is just the first step. Tailored video ads comes next. Sky provides the example again: Sources who have worked with Sky say that it is possible to serve nearly a million versions of the same ad, depending on who is watching and where. All of this happens in real time, using the TV’s set top box as a dynamic ad server. And it is possible today, in what we think of as linear, neutral television.
The opportunity Facebook and its peers see is twofold. First, they can offer even more sophisticated targeting tied in with a user’s online behavior. Second, smaller audiences mean lower prices than on television, a mass medium. In essence, tailored, targeted video ads combine the promise of data-rich internet advertising with the immersive quality of television.
Yet, targeting and personalisation remain imperfect even in text. But that’s not going to stop the giants of the web from filling the internet with video ads.