How Companies Use Data
The internet is full of free apps, free social networks, free search engines, and free streaming platforms. But building and running these services costs billions of dollars. If you are not paying with money, something else is being exchanged. In most cases, that something is your data. Understanding exactly how companies collect, analyze, and monetize personal data is one of the most important skills of the digital age.
Step One: Collection
Data collection begins the moment you interact with a digital service. Companies collect data in several ways. Direct collection happens when you fill out a form, upload a photo, or type a search query. You are consciously providing information. Passive collection happens in the background. When you use an app, it can record how long you spent on each screen, what you tapped, what you ignored, and how fast you scrolled past certain content. Websites run small programs called trackers that follow you across the internet — even to sites owned by completely different companies. Third-party data comes from data brokers — companies whose entire business is gathering information from many sources and selling the combined profile. They might buy your purchase history from a retail loyalty program, merge it with public records, layer on your social media activity, and sell the result to any business that wants it.
A tracker is a tiny piece of code embedded in a website or app. When you visit a page, the tracker fires a request to the tracker's server, telling it exactly which page you visited and when. Because the same tracker appears on millions of different websites, the tracker's owner can build a timeline of everywhere you have been online — even across sites you think have nothing to do with each other.
Step Two: Profiling
Raw data is just a pile of numbers. The valuable part comes from profiling — the process of analyzing data to build a model of who you are. A profile might estimate your age, your household income, your political views, your health status, your relationship status, your biggest anxieties, and what you are most likely to buy next. Profiles are built using statistics and machine learning algorithms that find patterns across millions of users. If people who searched for certain terms, liked certain posts, and visited certain stores at certain times tend to share a characteristic — the algorithm learns that pattern and applies it to new users who look similar. The result is that a company can often predict things about you that you have never explicitly told anyone.
Companies do not need you to state your income, health status, or emotional state for that information to enter your profile. It can be inferred from your behavior. One study found that a person's Facebook likes alone could predict personality traits more accurately than their own friends could. Inferred data carries the same power as stated data — and you often have even less awareness that it exists.
Step Three: Targeting and Prediction
Profiles are used to target you. Advertising targeting means showing you an ad specifically because of what your profile reveals about you — not because of what page you happen to be reading. If your profile suggests you are anxious about college admissions, you may see ads for tutoring services. If it suggests you have recently started running, you may see shoe ads. Prediction goes further. Companies use profiles to predict future behavior: what you will search for, what you will buy, whether you will cancel a subscription, and how you will vote. Platforms use behavioral predictions to decide which content to show you next — optimizing for engagement, which means keeping you on the platform as long as possible.
Match each business practice to what it actually means.
Terms
Definitions
Drag terms onto their definitions, or click a term then click a definition to match.
The Feedback Loop
There is a powerful feedback loop at work. A platform uses your behavioral data to predict what content will keep you engaged. It shows you that content. You engage with it. Your engagement creates more behavioral data. That data is used to refine the predictions. The loop runs continuously, optimizing for one goal: maximizing the time you spend on the platform. This loop is not designed with your interests in mind. It is designed to maximize advertising revenue. Sometimes those interests align — you find content you genuinely love. But the system is equally good at surfacing content that triggers anxiety, outrage, or compulsive checking, because those states also drive engagement.
Why do many popular apps and websites cost nothing to use?
What does behavioral targeting mean?
What is a data broker?
Follow the Data
- Step 1: Choose one app or website you use regularly. Look up its privacy policy — most can be found at the bottom of the website or in the app's settings.
- Step 2: Read the 'data we collect' or 'information we share' section. List five types of data the company says it collects.
- Step 3: For each type of data you listed, write one sentence explaining how that data could be used to build a profile of you.
- Step 4: Does the company say it shares data with third parties or data brokers? If yes, what does it say about that? If no, does that surprise you?
- Step 5: In two sentences, describe what surprised you most about what this company collects and why.