Explained: Here’s How Advertising Tracks You Across the Web
At last week’s Family Online Safety Institute conference, the message was largely optimistic. Parents feel their kids are safer than ever online. Still, three-quarters of parents worry about the information Web marketers collect about their kids, according to a new FOSI survey. Another study released last week by the Pew Research Internet Project found that nine out of 10 adults feel they’ve lost control over how their personal information is being used online.
Parents are more concerned about online ad networks than they are about screen-time addiction or creepy dudes on Twitter hitting on their teenage daughters. On the parental worry radar, online tracking is right up there with pornography and cyberbullying.
Parents should be concerned about the information marketers are gathering, about both their children and themselves. But a lot of the worry is free-floating; most people don’t know what “online tracking” actually is, or they think it’s something it isn’t.
Who can blame them? Just check out the migraine-inducing “LumaScape” chart that illustrates the insanely complicated ecosystem behind online ads.
In this column I will explain how Web ads work, what advertisers know about you, and how that might change.
If you are raising children of the Internet, pay attention. What your kids do online could influence the college they get to attend, where they work, whether they qualify for loans or insurance, and a lot more.
When you visit a website that contains ads, those ads leave behind text files on your computer called “tracking cookies.” (Don’t ask why, just go with it.) These files contain a combination of letters and numbers — kind of like a license plate on steroids — that identify your particular browser on your particular computer.
Tracking cookies do not, by themselves, identify you. They ID a browser on a particular device. (If you open a page in Chrome and then load the same page in Firefox or Safari, or open the same site on a different computer, you will get new cookies with different IDs inside.)
When you visit a new webpage, the machines that deliver ads check your cookies, find your unique ID, and record information about your online behavior. They might also deliver ads based on your apparent interests. For example, if you spend a lot of time at Edmunds.com, you’re more likely to see ads for the new Volkswagen Golf no matter where you go on the Web. If you also spend time at Bleacher Report and Robb Report, the networks will assume you’re a high-income male of a certain age and target ads accordingly.
The vast majority of ad networks do not keep a record of every site you’ve visited (Google being a major exception). They simply use the sites you visit to make broad conclusions about products you might be interested in. Then they discard the URLs.
(As my colleague Rob Pegoraro has pointed out, Verizon and AT&T mobile networks identify you in a different way — by putting a unique ID on the webpage request of every page you visit. Unlike cookies, which you can delete and manage, this ID is pretty much impossible to get rid of. Last week AT&T announced it would stop using these IDs. There’s also something called “browser fingerprinting,” which uses your software settings to identify you, but that’s a topic for another day.)
Whether it’s called interest-based, targeted, or online behavioral advertising, it all means the same thing: The ads you see are personalized based on what you do online. Why do Web publishers allow these ads on their sites? In a word: money. Targeted ads are worth roughly twice as much as non-targeted ones. Much of that complicated ad ecosystem is built around matching relevant ads to the right people in real time.
You can use browser plug-ins like AdblockPlus or AdMuncher to nuke ads. But please don’t, because online advertising pays the salaries of almost all the journalists writing the stories you’re reading. Like this one.
You are what you click — or not
Here’s the funny thing. The information these ads use to target you is often wildly inaccurate and downright weird. You may be a NASCAR dad, but the advertising network thinks you’re a soccer mom. You may be renting a trailer from your half-wit cousin, but it thinks you own a home in Westchester County.
There are a handful of ways you can find out exactly what kind of person advertisers think you are and tell them to quit sending targeted ads.
Option One is to click on the teensy-tiny Ad Choices icon that appears on more than 80 percent of all targeted ads, like this one from the Yahoo.com home page.
Clicking that link takes you to a page with information about the ad and how to opt out. Choose the “manage” link to select different categories of ads you do or don’t want to see.
That in turn shows you the “interest categories” you fall into. From here you can select which types of ads you want to get, or opt out of all interest-based ads with one click. (If you’ve already opted out, you won’t see any interests displayed.) You will still see ads; they just won’t be personalized.
This is not unique to Yahoo. Targeted ads on other advertising networks also let you manage the types of products being pitched or opt out entirely.
Option Two: Go directly to the sites of the biggest data collectors, such as BlueKai, eXelate, or Lotame. These profiles are usually much more detailed, in part because they often combine online activity with real-world demographic data — like your age, income, marital status, and more — collected by companies like Nielsen and Acxiom. Your name and other personal information are stripped out before they deliver ads.
You can choose to opt out of most targeted ads by using the Digital Advertising Alliance’s consumer choice page, but this doesn’t accomplish much in practical terms, and it’s an enormous hassle. You’ll have to do it on every machine and every browser you use, and even then ad companies will still collect your profile data — they just won’t display ads based on it.
If you don’t want to be tracked, your best and easiest option is to install a browser plug-in like Ghostery or Disconnect, which lets you choose which if any ad networks you want to profile you. Then get ready to see a lot of cheap ads for reducing belly fat or for cut-rate loans, since the high-priced, targeted ads will no longer be able to find you.
So your Web surfing history is gathered anonymously to deliver ads, based on profiles that are often comically inaccurate. What’s to worry about?
Today, not so much. But as online tracking data gets combined with other information and run through inscrutable algorithms, that’s likely to change very soon.
Facebook, which began using Web tracking data to deliver ads earlier this year, is a good example of this. Click on the hidden X in the upper-right corner of a Facebook ad and you’ll see a pop-up menu offering to hide the ad, like it, or find out more about it.
Click Why am I seeing this? and up pops a new window with more information. Select Manage Your Ad Preferences from that window and you come to a third page listing the different categories of ads Facebook thinks you might be interested in seeing. For me, Facebook lists nearly 600 categories. These are a combination of places I’ve visited outside Facebook and things I’ve liked inside Facebook, but it also includes things Facebook assumes I would like, based on the data it has on me.
Most of the preferences are spot on; some (like the above) are wildly off base. And, accurate or not, these assumptions could come back to bite you.
If you say you like bikinis and beer, for example, and data shows that people who like bikinis and beer tend to be poor credit risks, then your online application for a loan could be rejected, despite your actual creditworthiness. If your pattern of likes and Web visits correlates well with those of people who smoke, you might receive a higher quote for health insurance, even if your lips have never touched a cigarette.
This is not just theoretical. In 2011, the Digital Advertising Alliance adopted a set of voluntary principles that forbid member companies from using tracking information “when making adverse decisions with respect to employment, credit, healthcare treatment, or insurance eligibility.” In other words, it knew exactly how the data it was collecting could be used.
That the ad industry took this step is awesome, but the key word in that sentence is voluntary. Most major Internet ad companies have agreed to these principles (including Yahoo), but not all. And if an ad company decides to go rogue and sell this information to third parties, it’s not clear what if anything will happen to them.
Does this mean you should toss your laptop and smartphone into a limestone quarry and swear off the Internet? No. Just remember that anything you do online may have consequences offline. As in the real world, your behaviors on the Web affect your future.
Questions, complaints, kudos? Email Dan Tynan at ModFamily1@yahoo.com.