It’s not just Facebook — online quizzes also collect data about you
Online quizzes help us understand ourselves: from what kind of ‘90s kid we were to which fairy tale was actually about our life. But more often than not, each question gives the site you’re on (and the companies they share your answers with) valuable insight into your interests.
The privacy trade-offs that come with using social media are now clear to many. On Facebook, for example, almost any information you give the service is sold to marketers in the name of targeted advertising. While this has been true for more than a decade, the recent Cambridge Analytica scandal has made it more clear than ever. But internet users are tracked all over the web even when they’re not on Facebook.
“Any survey asking you anything, from, ‘What would you do in a certain situation?’ to your favorite color, is gleaning information about you that is useful to algorithms,” Max Kilger, the director of analytics at University of Texas at San Antonio, said in an interview. “Any quiz you take, on or off of Facebook, has a chance that the site will resell it or combine it with some existing data on you.”
Cambridge Analytica made use of a personality quiz called “thisismydigitallife.” 270,000 Facebook users chose to take the quiz, but the test required takers to give access not just to their profile data but the profile data of all their Facebook friends as well. In sum, 50 million people had their personal info sent to Cambridge Analytica, most of whom hadn’t taken the test. According to Facebook, this was not a breach of its service.
There are other large-scale quizzes like this around the internet, but simple ones siphon off data, too. “That is a classic use of quizzes and personality tests,” Pam Dixon, executive director of World Privacy Forum, said in a phone interview. “With most quizzes online, it’s important to question if this quiz exists just to grab my personal data, likes and dislikes, in order to further target my information” Dixon said. “A medical themed quiz, for example, may ask for symptoms. Unless it’s your hospital or doctor presenting you the quiz, it’s usually not covered under any health privacy law.”
Here are some examples.
Google Surveys’ privacy report can be found here. While the search engine states that it doesn’t sell your data to outside advertisers, those who use the Google Survey platform have the ability to learn granular details about their users who take the survey. Sites can even get paid by Google if users take part in surveys, thanks to the company’s Opinion Rewards tool.
The news and entertainment site BuzzFeed, arguably, popularized internet quizzes. Quintessential questions like “Which Emoji Am I?” or “Can You Cook Dinner For Two On A Budget?” are answered via a simple line of questioning. Though it’s easy to see how certain questions, when shared with the marketers, give companies insight on what you really like. This offers companies a better idea of who you are and what the best ways are to convince you to buy things.
“You know which muppet you are, but BuzzFeed doesn’t attach your quiz results to personal information,” a BuzzFeed spokesperson said. “All quiz response data is anonymized. We don’t sell our quiz data to clients. Any data shared with clients is aggregated, anonymized, and stems solely from their own campaigns,” the spokesperson said.
The company uses “the aggregate information to make better quizzes and content, both editorial and branded,” they said. That information is not used “to change what people see on the site, but we have discussed whether the Rowlfs of the world would prefer to see certain types of content and how we could potentially personalize content in a transparent manner,” the spokesperson said.
While information gleaned from regular BuzzFeed quizzes stays within the company, data from branded editorial quizzes is shared with the company sponsoring the post (e.g.: Verizon in this posts’ case). BuzzFeed’s quizzes run off of Google’s Survey platform.
Dan Barker, an e-commerce and analytics consultant, was one of the people engaging in this conversation back in 2014. His previous blog post on the matter pulls back the curtain to show what really goes on behind a quiz. “What BuzzFeed does isn’t unusual — in fact their approach to data seems to be friendlier than many other similar sites,” Barker said in a Twitter message. “There are lots of poor practices out there which end users seem unaware of, and this is at the more benign end of the scale. Hopefully awareness will change in response to greater coverage of the issues.”
Barker says people who are dying to take part in a quiz should use incognito mode or a private window on their browser.
BuzzFeed keeps all the information it learns about you to itself and it doesn’t affect what you see around their site (for now). Other sites, however, that utilize Google Surveys as well as places like Yougov, Pollfish, Quicksurveys and even Cambridge Analytica’s questionnaire can be used to build profiles on their users.
Who cares if they have a bunch of useless quiz data?
Tech companies continue to understate or hide in fine print just how much data they really take from users. Many might respond, “Who cares?” While some may hate the idea that companies are recording quiz answers to build a picture of you and share with other unnamed companies, others could not care less. If you have nothing to hide, why should you care if anyone can see how you’re answering some dumb quizzes?
“We’ve seen how machine learning and algorithms can group people into buckets,” Dr. Kilger said. “Very innocuous questions can lead to many decisions being made behind users’ backs. They may not be aware of what they’re missing out on all because an algorithm somewhere said ‘well, this isn’t a very good prospect.’” A real-world example of this surfaced when Pro Publica revealed Facebook allowed housing ads to exclude those categorized as African-American, Spanish-speakers, mothers of high school kids and more.
The potential problems go beyond housing. “That health data you gave that quiz, for example, could result in you being charged more for health insurance or can determine what lead generation you get around credit,” Dixon said. “Your data matters and it lasts a very long time.”
April 2, 2018, 12:30 p.m.: This article has been updated.