The way that sites like Facebook can anticipate our interests can feel quite uncanny, especially if ads pop up for something we’ve thought about but haven’t searched for.
Despite lots of conspiracy theories out there about tech giants listening in on your life, there is no evidence today that Facebook (or Google or Apple) is actively listening to the microphones of our computers and mobile phones.
Sure, there have been reports of smart televisions or other internet-connected 'things’ at home being used by the CIA or MI5 to spy on people, and there have been warnings that what we say in front of the TV could be recorded and distributed to a third party.
This isn’t a general trend, however. There are many other ways data-driven companies find out about our interests and anticipate our mood and needs, and in short know a lot about our personal life.
Media platforms collect so much information about us, that sometimes they can predict what we might be into before we even know it ourselves - and then they target us with ads they think we will respond to.
People are often astonished and uncomfortable about how accurate this targeting can be. Internet advertising is broken into various categories, from random ads to highly targeted ones. The most expensive - and of course the most effective or relevant ads - will be those that are the closest to our current interests or to our mood at a particular time and location.
Internet advertising platforms don’t have to listen to your conversations to do that. You are tracked everywhere you go on the web.
People who use social media platforms give them lots of information about themselves, such as ‘liking’ the fan page for a particular band, or filling in a quiz. Over time and en masse, so much data builds up collectively that this data can be used to infer an implicit data set – for example, inferring that most people who ‘liked’ a band could also have a certain personality trait. This information is frequently on-sold to tracking entities like advertisers or analytics firms.
In a 2014 study of users of Facebook and Chinese social media site ReRen, we found that almost a quarter of Facebook apps provide user’s information to one or more of these fourth-party tracking entities. In other research in 2012, a small group of volunteers allowed us to analyse some of the seemingly harmless information they left on Facebook, such as music-related likes, and we were able to infer surprisingly accurate private information including their socio demographics and daily habits.
Almost all commonly used pages on the web are embedded with invisible tracking components that follow you wherever you go on the web and send information about you to Google, Facebook, etc. That’s why you will look at something on eBay for example, then find it popping up when you go to a news website. Mobile apps are no different.
Ads follow you through a technical process called real-time bidding – they ‘talk’ to Google milliseconds after you visit a page, and advertisers compete with each other to show their ads to you, whether they are targeted (matching your profile the closest), re-marketing (matching some earlier interest you had) or contextual or location-based, etc.
That’s why they can be spookily accurate about your taste in holidays or office accessories.
While building your profile, Facebook and others don’t just stop at your web history. In essence, everything you contribute is good to take, ranging from your emails (probably scanned by analytics algorithms), to your comments on a political post or a simple angry face as a reaction to one of your friend’s posts.
Better privacy tools are being developed but until then, you can protect your own privacy by being careful about the information you share about yourself online, particularly via free platforms like Facebook. As the saying goes – if you aren’t paying for the product, then most likely – you’re the product.