How Much Data is Facebook Tracking?
The scale of social media data collection — from Facebook tracking to modern AI data harvesting.
title: "How Much Data is Facebook Tracking?" slug: "how-much-data-facebook-tracking" description: "The scale of social media data collection — from Facebook tracking to modern AI data harvesting." datePublished: "2013-01-20" dateModified: "2026-03-15" category: "AI & Privacy" tags: ["Facebook", "tracking", "privacy", "data collection"] tier: 3 originalUrl: "http://www.applieddatalabs.com/content/how-much-data-facebook-tracking" waybackUrl: "https://web.archive.org/web/20130120093934/http://www.applieddatalabs.com:80/content/how-much-data-facebook-tracking"
How Much Data is Facebook Tracking?
When I wrote this piece in early 2013, the answer to the title question was "a lot more than you think." We documented how Facebook tracked not just your posts and likes, but your browsing activity across the web through embedded Like buttons and Facebook Connect logins. We estimated that Facebook was collecting data on every user across hundreds of data points. That number felt alarming at the time.
Today, the company that was once called Facebook has rebranded to Meta, absorbed Instagram and WhatsApp, launched Threads, built a VR platform, and assembled what might be the most comprehensive personal surveillance infrastructure in human history. The 2013 version of this article was describing a garden hose. The 2026 version has to describe the ocean.
What Facebook Tracked in 2013
Our original article broke down Facebook's data collection into categories. There was the obvious stuff: your profile information, posts, photos, check-ins, friend list, messages, and likes. Then there was the less obvious stuff. Facebook's Like button, embedded on millions of websites, functioned as a tracking pixel even if you never clicked it. Simply loading a page with a Like button sent your browser cookies back to Facebook, telling the company what sites you visited, when, and for how long.
Facebook Login, the feature that let you sign into third-party apps and websites using your Facebook credentials, gave the company a view into your activity across those services. Facebook Exchange, launched in 2012, let advertisers retarget you based on your browsing history. And Facebook's mobile app, which by 2013 was the primary way most users accessed the platform, collected device information, location data, contact lists, and call logs.
We tried to count the specific data points. Profile fields, behavioral signals, inferred interests, social graph connections. The number was already in the hundreds. Most users had no idea how much was being collected because the privacy settings were deliberately complex and the data disclosures were buried in a Terms of Service that would take about 75 minutes to read.
In 2013, Facebook tracked hundreds of data points per user. After acquiring Instagram, WhatsApp, and building a VR platform, Meta collects data from nearly every form of human communication except in-person conversations. Give them time.
The Empire Expanded
Between 2013 and 2026, Facebook's data collection grew in three directions: more apps, more data types, and more users.
Instagram, acquired in 2012, grew to over 2 billion monthly active users by 2025. WhatsApp, acquired in 2014 for $19 billion, has 2.7 billion users and is the primary messaging platform in much of the world. Despite WhatsApp's end-to-end encryption (messages are encrypted in transit), Meta collects metadata: who you message, when, how often, your contacts, your profile photo, your status updates, and your location data. In January 2021, WhatsApp updated its privacy policy to allow sharing more user metadata with Facebook for ad targeting, triggering a user revolt that drove millions to Signal and Telegram.
Threads, launched in July 2023 as a Twitter competitor, hit 100 million signups in five days, largely because it piggybacked on Instagram accounts, pre-populating user profiles and social connections. It was a masterclass in data-driven growth: Meta used existing Instagram data to make Threads instantly useful, which also meant Threads users were adding yet another data stream to their Meta profile.
Then there's the Meta Quest VR headsets. The Quest 3, released in 2023, tracks head movements, hand gestures, eye direction, room layout (via passthrough cameras), and voice commands. Meta requires a Meta account to use the device. The privacy implications of a VR headset that maps your living room and tracks where you look are significant, and Meta's privacy policies give the company broad rights to use this data for product improvement and advertising.
The Regulatory Reckoning
The EU hit Meta with its largest GDPR fine ever: $1.3 billion (EUR 1.2 billion) in May 2023 for transferring European users' data to U.S. servers without adequate privacy protections. It was the culmination of a decade-long legal battle initiated by Austrian privacy activist Max Schrems. Meta had already been fined $414 million by the Irish Data Protection Commission for its Instagram practices and $400 million for failing to protect children's data.
Apple's App Tracking Transparency (ATT), introduced with iOS 14.5 in April 2021, let users opt out of cross-app tracking with a single tap. Roughly 75% of users opted out. Meta estimated the change would cost it $10 billion in advertising revenue in 2022 alone. The company's stock dropped 26% in a single day in February 2022, erasing over $230 billion in market value.
Meta's response was to invest heavily in AI that could predict user interests from on-platform behavior, reducing its dependence on cross-app tracking. The company's Advantage+ advertising system uses machine learning to optimize ad targeting and creative without relying on the third-party data that ATT blocked. It worked. By 2024, Meta's ad revenue had recovered and set new records.
What This Means for Organizations
Meta's data story matters for any organization because it illustrates both the power and the risk of large-scale data collection. The company built a $1.5 trillion business on user data, but regulatory backlash, privacy changes, and public scrutiny cost it tens of billions and required a fundamental restructuring of its ad technology.
Organizations building their own data collection practices should learn from Meta's experience. Data governance isn't optional when regulators can issue billion-dollar fines. And data strategy needs to account for the possibility that your primary data sources could be disrupted overnight, as Apple disrupted Meta's cross-app tracking. Building resilient, operational AI systems means planning for a world where data access is constantly shifting.