Today, instead of talking about one specific ad, I want to talk about online advertising. You’re reading this online, so you probably spend time browsing the internet, which means you likely see a fair amount of online advertising. The thing about online advertising that’s interesting is that it’s entirely unlike advertising anywhere else in that different viewers can visit the same sites and see different ads. When you watch a television show or pick up a magazine, you’re seeing the same ads every other person who watches that show or reads that magazine sees. When you drive down the highway you’re seeing the same billboards every other driver on that highway sees. But when you visit, say, your favorite news website today and read a story, you might see different ads from what other people reading the same article on the same site at the same time are seeing.
There can be a couple of different reason this happens, but increasingly one of the most popular with advertisers is because the ads are “targeted” to you. This means that they are showing you ads specifically chosen based on what they know about you that they believe are most likely to interest or appeal to you. What they think will appeal to you is all based on demographics and stereotypes. Advertising on Facebook is designed to be entirely targeted advertising like this. So, for example, I’m a woman in my twenties so I see a lot of ads for weight loss, bikini waxes and a Facebook game aimed at young women. My husband, on the other hand, sees ads for AT&T, tech degrees and fantasy games. And while I frequently see ads mentioning moms, he only rarely does.
It’s pretty easy to see how targeting advertising works on Facebook (especially considering it’s tied so closely to the mess they’ve made recently of their privacy practices), but the process by which it works on the rest of the internet is far less transparent. There are various companies that specialize in targeted advertising and each gathers information about you as you browse the internet in order to display what it considers to be the most appropriate ads for you whenever you visit one of the websites it works with.
One such company is AudienceScience. I’m specifically mentioning them because they have a particularly interesting website with information on what they think certain groups of people are most likely to be interested in. They do not, sadly, have much information about how they determine which groups to put people into, however. For example, I appear to be in one of their car groups, based on the ads I see on their partner sites, and I have no idea why that would be, since I have no interest in cars and can’t even remember the last time I visited a car website on this computer. To check out their groupings and the interesting information they list about each, visit their periodic table-like visualization and click on any group or tab at the top you want more data about.
AudienceScience also has a video explaining in vague, simplistic, somewhat unhelpful terms how their system works. It’s clearly designed to try and convince people that advertising is good for them (something much of the advertising world wants to convince us of and I’m rather skeptical about, given the irresponsible nature of so many of the ads I see), but it is interesting.
So does this supposedly very scientific method of targeting ads to viewers tastes work? Well, I think that probably depends on the viewer. For me, it doesn’t seem to. As I already mentioned, I see a lot of car ads which I really can’t figure out. Not only am I pretty much not at all interested in or attracted to cars (I hate driving and find cars fairly boring), but I also just recently bought a car and have absolutely no plans to buy a new one for… hopefully ten years at least.
Here’s a sampling of some of the ads I see on websites affiliated with AudienceScience, so they are presumably targeted to me using their method (I’ve included only one of any given type).
This sampling doesn’t really reassure me that AudienceScience’s method is as innocent or transparent as they’d like us to believe it is. For example, the ad for “Austin 1-Day Deals” may only mean they know I’m in Austin, but I have to wonder what in my browsing history would tell them that, since I rarely visit any websites about Austin or local attractions/businesses and visit websites for local things in the city I used to live in and places I have friends or family living far more often. And while I’m guessing they got that information from my IP address, I’m also not sure I can agree with them that such information doesn’t fall under “personally identifying information”, since knowing my location is pretty personally identifying.
So while I think the idea of ad targeting might have some merit, I don’t think it’s done very well right now. I think that advertisers need to work on being more transparent about how they gather information and use it as well as simply what information they have. I think that the assumptions they make once they have that information need to be more well considered. Being female and in your twenties does not automatically mean you have any interest in Sex in the City, Twilight, weight loss, beauty products or pink electronics. Being male and in college doesn’t mean you play videogames, are into sports, want to drool over mostly naked chicks or drink beer. There should be ways to target advertising without collecting any personal information at all and without making so many assumptions about a person’s interests.