After toggling off all seven of Facebook’s ad tracking and information settings, I resumed my News Feed scroll expecting to be regaled with ads for beef jerky and car insurance. Instead, I realized just how much Facebook knows about me and scrolled past ads that, instead of becoming less relevant, only felt more personalized.
Facebook knows that I’m pregnant. Facebook knows that I’m a photographer. Facebook even knows that I’m currently remodelling my bathroom.
But the bigger problem is not what Facebook knows, but what Facebook users don’t know. If a tech writer doesn’t know that turning off Facebook ad settings only diverts ad tracking to the data tracking that is impossible — or at least time consuming — to turn off, what does the rest of the network’s two billion users really understand about what’s happening to their data?
If MIT graduate in privacy-preserving machine learning and staff technologist for the Electronic Frontier Foundation (EFF) Bennett Cyphers calls Facebook data tracking “opaque and difficult to understand,” how can the average user grasp what’s happening on the biggest social media network?
A PEW Research Study suggests that three-quarters of Facebook users are clueless about how Facebook determines what ads they see. “There are several misconceptions on Facebook ads,” Cyphers told Digital Trends. “One, that you can opt-out of targeted ads. Two, that you can opt-out to protect your privacy when Facebook still collects the exact same amount of data.”
Advertisers have always aimed to reach specific audiences that may be more receptive to a certain message — it’s why you see commercials for disability lawyers during daytime television and commercials for toys during cartoons. With web cookies and social media, however, online advertisers have more options for tailoring that message than ever before.
Facebook’s Ad controls allow users to turn off ad data in three categories, a simple toggle that’s available in the ads section of the settings. The first toggle turns off data based on partners on user activity off of Facebook. This turns off Facebook Pixel, a data tracker that website owners install to track visitor data. Facebook Pixel is the technology that’s responsible for you seeing an ad for the exact same pair of shoes that you added to your cart in that online store a few hours ago.
What you click on, hover over, or even search for on Facebook Marketplace may be used to target your future ads.
Facebook ads don’t exist solely on Facebook-owned platforms — the second toggle revokes permission to have Facebook use your data on the ads that you see outside of Facebook, which can pop up anywhere from websites to apps. The company is responsible for a surprising number of ads that exist off of Facebook and Facebook-owned platforms like Instagram and WhatsApp.
The final toggle in the ad settings turns off the setting that would inform your friends that you like a certain Page when displaying an ad for that Page.
Along with those three settings, Facebook allows users to toggle off ads based on the information in the about section, using the “your information” section of the ad controls. This option turns off ads based on the profile fields for relationship status, employer, job title and education — it’s where you go if you’re annoyed that, as soon as you changed your relationship status to married, you started getting ads for diapers and baby formula — before even coming home from the honeymoon.
But the Facebook ad settings that you can turn off with a simple toggle are only part of the complex equation that determines what ads you see woven in between memes and pictures of what your best friend ate for lunch. The “Your Interests” section of the settings page tracks what Facebook thinks you’re into based on the posts that you like, links that you click on, or even posts that you hover over for longer than usual.
The interests section is largely why, after switching off the toggled ad settings, my own Facebook ads felt more personal — because the posts I tend to click on say a lot more about me than my relationship status or education. While the interests section isn’t impossible to turn off, it’s time-consuming to disable each individual interest one by one. There’s no option to delete all the interests Facebook has gathered on you at once — you’ve got to click the X every single time.
I contacted Facebook to see if they could explain why it’s so difficult to disable your ad preferences, and why there’s no batch-delete option, but they’ve yet to respond.
Facebook still collects the same amount of data with the ad settings toggled off — it just doesn’t use as much of that data.
Facebook says that, once you remove an interest, that same interest won’t be added back if you continue clicking on similar posts. But while the network won’t add identical interests back, deleting one won’t prevent the network from adding similar interests. For example, I can delete the interest for parenting, but if I click on a link to Parenting Magazine, Parenting Magazine is added to my interests. What you click on, hover over, or even search for on Facebook Marketplace may be used to target your future ads.
The second major source of ad data that users can’t simply tap on a toggle to turn off is what’s called data brokering. While Facebook knows a lot about you, other businesses may already have some of your information. Hulu, for example, knows what you like to binge watch and eBay knows what you are looking to buy. Companies can upload lists of customer information to Facebook, using data like your email address or phone number to connect that information to your profile.
Along with companies that have your email because you’re a subscriber, data brokers — companies that exist solely to track information for targeted advertising — can also upload lists with your information and sell that information to advertisers. These lists can also be used to target a Facebook ad. If you sign up for email newsletters for a pregnancy blog or make a purchase from an online maternity store, your email may now be a part of a data list that’s uploaded to Facebook and sold to other companies, sending you pregnancy-related ads even if you haven’t clicked on anything telltale on Facebook itself yet.
Like with Interests, Facebook doesn’t offer an easy toggle setting to turn off that data. Advertisers that used one of those data lists can be deleted one-at-a-time on Facebook inside the “Advertisers and Businesses” section of Ad preferences by clicking the X on each individual company with your data. Or, users can opt-out from the data broker itself, one at a time, from a list that contains more than 50 data brokers.
While understanding what Facebook tracks and the limited settings that turn off some of that data use is a start, users should also understand that toggling a setting off does nothing to prevent tracking of that data — it just keeps Facebook from using that information to send targeted ads. Facebook still collects the same amount of data with the ad settings toggled off — it just doesn’t use as much of that data.
To be clear, none of the information that Facebook knows about me is exactly a secret, nor did I really expect that information to remain private once shared online. You can scroll through my public Instagram page and find a pregnancy announcement. Click on my writer’s bio, and you’ll see that I’m a photographer, find the state where I live, and learn what I like to do in my free time. But seeing ultra-personalized ads unnerves many Facebook users and begs the question, what, exactly, does Facebook know about me — and what, in turn, do Facebook advertisers know about me?
Because I also own a small photography business (another thing Facebook knows about me), I’ve been on both sides of a Facebook ad. And while seeing hyper-personalized ads is unnerving, I know that Facebook didn’t deliver a list of users with a specific interest that had my name on it to some random advertiser.
When boosting a post on Facebook, I can choose a geographic region, age, and gender of who that post will be boosted to. I can select from a list of different interests and profile information — for example, I can boost a post about my wedding photography specifically to Facebook users that list engaged as their relationship status.
I never see a list of names or users. I can guess that the users that liked that post may have been part of that audience, or may have seen that post through a share or a friend that liked or commented on the post. But, Facebook doesn’t explicitly tell advertisers who is part of that audience.
“People tell us they want to see ads that interest them most …”
“When people target ads on Facebook, Facebook isn’t sharing any data with them. That’s true in a sense, they don’t sell lists of data,” Cyphers said. “But people have shown it time and again that you can use targeting tools to extract data about people…if you target your ad in a certain way, you can know that whoever clicks on that ad is part of that targeted demographic.”
Facebook says it doesn’t sell user information — and while the network has been under increasing scrutiny since a third-party app sold user data in the Cambridge Analytica scandal, no one has found anything to the contrary. While Facebook hasn’t maliciously sold user data, that doesn’t mean the network is doing everything possible to protect user data. Even CEO Mark Zuckerberg himself admitted that Facebook didn’t “take a broad enough view” of its responsibility and allowed controls that gave access to third-party apps without user permission. (Like with Cambridge Analytica, third-party apps may not fall under the same restrictions, particularly those quiz apps that are actually designed to gather your information when you click to allow access to your data.)
“People tell us they want to see ads that interest them most, which is why we encourage people to use the tools available to understand and control how their information influences the ads they see,” a Facebook representative told me. “We believe relevant advertising and privacy aren’t at odds and are committed to doing both well.”
Facebook may know a lot about its two billion users in order to provide an anonymous way for advertisers to reach the people that may actually be interested in that ad, but the problem with ad data isn’t one of anonymity.
The issues with Facebook ad data are threefold — first, the complexity that makes understanding and managing that data difficult for the average user, second, the possibility of discrimination in ad campaigns, and third, the ability for a few bad players to use what’s largely a helpful tool for businesses to create targeted misinformation campaigns and propaganda. Facebook has at least taken a few small steps to address each one — adding new tools to help users understand ads, restricting the targeting tools for housing and employment ads, and adding new restrictions for political advertisements — but there’s arguably much more to be done.
Facebook earned more than $16 billion in ad revenue for the first three months of 2019.
“I don’t think data being anonymous is the problem,” Cyphers said. “First of all, anonymous means different things to everyone you ask. There are ways that shady actors can use tactics to get information. I don’t think that’s the problem. The problem is using it to discriminate. There are a lot of ways for shady advertisers to use that and it’s hard to police.”
Facebook recently made a number of changes after a lawsuit from the Department of Housing and Urban Development over the platform’s allowed use of demographic information such as race to target housing ads. The network has faced similar criticism for job ads.
The way Facebook is set up, Cyphers added, makes it difficult to audit campaigns that may be discriminatory or may, in fact, be targeted misinformation. For the most part, it’s up to Facebook to audit those campaigns.
The same targeting tools that can help small businesses succeed can be manipulated to target misinformation to the users most likely to believe them, Cyphers says. Among the list of interests that Facebook tracks, the social network uses labels like political party affiliation as options that can be used to target ads, for example. The interest section also includes an entire category of “news and entertainment” that contains the publications that a user follows. That would theoretically allow bad actors to target a post towards a liberal or conservative that also tends to like posts from an outlet with a penchant for fake news.
Ads aren’t going away — Facebook earned more than $16 billion in ad revenue for the first three months of 2019. And while you could always quit Facebook, that’s not an option for many people — or many businesses.
Getting rid of ads entirely would stunt the businesses that rely on the affordable but relevant advertising that Facebook and similar outlets provide, and in turn, could affect the jobs that the company is able to offer. And while Facebook is one of the biggest players, the social network isn’t the only one using user data to sell ads online.
The issue, however, is larger than what individual Facebook users can tackle on their own. Legislation to make data privacy a priority for the social network — and in turn simplifying the privacy settings — would be a big step forward, Cyphers suggests. Even just allowing you to delete your interests or ad data in batches as opposed to one at a time could be a major help.
While much of the responsibility lies with the company and the organizations working to protect user privacy, that doesn’t mean there’s nothing for users to do. Turning off the available ad tracking options may not lead to completely irrelevant ads, but it helps. And while eliminating all the topics in the interests section would likely take a few hundred clicks, users can turn off the interests more likely to be abused in a misinformation campaign, such as labels for political affiliations, and the pages and organizations that may lean towards a specific viewpoint. The diligent can also go through the opt-out process for each of the data brokers mining data off Facebook as well.
Above all, Cyphers says, users should be in tune to the likelihood of misinformation campaigns and aware of their own biases. Messages that create a big emotional response are much more likely to be part of a misinformation campaign. “Educate yourselves and think about your own biases,” he said. “When you see an ad that really inflames the senses or a message that really strongly appeals to you in an emotional way, think about who is trying to reach you with that message and why.”
- How iOS 14’s privacy features keep you safe
- The best iPad Pro games
- The 100 best Android apps (August 2021)
- The best web browsers for 2021
- The best free movie apps for Android and iOS