Surveillance Capitalism in Plain English

Surveillance capitalism is the name for an economic model in which your everyday behaviour — where you go, what you tap, how long you hesitate, who you message — is quietly recorded, turned into data, and used to predict and nudge what you will do next. Those predictions are then packaged and sold to whoever wants to influence you. The surveillance is not a bug or an overreach on the edge of an otherwise ordinary business. It is the business. Once you see that, a great deal of what feels strange about modern technology — the sense of being watched, the ads that arrive before you have spoken a word — stops looking like a series of accidents and starts looking like a design.

The phrase belongs to the scholar Shoshana Zuboff, who set it out in her book The Age of Surveillance Capitalism. Her achievement was to give a precise name to something most of us only felt vaguely: that the leading technology companies had discovered a new kind of raw material and a new way to profit from it, and that the rest of us had never agreed to any of it. This piece explains that idea in plain English — what is taken, how it is refined, who buys the result, and why it is so hard to opt out.

What is actually being taken

Start with the thing itself. When you use a search engine, a map, a social app, or a smart speaker, you hand over some information simply to make the service work. A map needs your location to give directions. A search box needs your query to return results. That is a fair exchange, and if the story ended there we would have nothing to talk about.

But the story does not end there. Alongside the data a service genuinely needs, these systems collect a vast amount more: the route you did not take, the search you typed and then deleted, the photo you lingered on, the hour of night you tend to reach for your phone, the tremor in how fast you type when you are anxious. None of that is required to give you directions or search results. It is captured anyway, because it is valuable for something else entirely.

Zuboff calls this extra material behavioural surplus — the surplus being everything harvested beyond what the service needs to function. It is the exhaust of your ordinary life, the by-product you shed without noticing. And the central insight of surveillance capitalism is that this exhaust is not waste. Pooled across billions of people and fed into predictive models, it becomes the most valuable input the company has.

The data a service needs to work is the smaller half of what it takes. The surplus — the exhaust of your ordinary life — is where the money is.

From behaviour to prediction to sale

Here is the whole arrangement, laid out step by step, because the steps are where the meaning lives:

1. Extraction. Your behaviour is recorded — inside the app, and often far beyond it, through trackers embedded in other websites, other apps, and the physical world.
2. Surplus. Far more is taken than the service requires. The excess is behavioural surplus, and it is claimed as the company’s free raw material.
3. Refinement. The surplus is fed into machine-learning models that look for patterns — profiles of what people like you will click, buy, believe, or ignore.
4. Prediction. The models output forecasts about your future behaviour. These forecasts are the actual product.
5. Sale. The predictions are sold, not to you, but to parties who want to act on them — chiefly to influence what you do next.

Notice where you sit in this chain. You are not the customer. You are the source of the raw material at one end and, at the other end, the target the prediction is aimed at. The customer is a third party — an advertiser, a political campaign, an insurer, anyone willing to pay for a more reliable guess about what you are going to do, and for the chance to tilt it.

Zuboff makes a sharp point about the direction of travel here. The early versions of this business simply predicted your behaviour. But a prediction becomes far more valuable if you can also shape the behaviour to make the prediction come true. So the models do not only forecast; they are wired into systems that intervene — a nudge, a prompt, a perfectly timed offer — designed to steer you toward the predicted outcome. The product being sold shifts, subtly, from knowing what you will do to arranging that you do it.

How this differs from ordinary advertising

A reasonable objection at this point is that advertising has always tried to influence us. A billboard, a television spot, a newspaper ad — all of them attempt to change behaviour, and nobody calls a billboard surveillance. So what is genuinely new?

The difference is in the raw material and the direction of the gaze. Old advertising was broadcast: the same message went out to everyone who passed the billboard or watched the show, and the advertiser knew you only as a rough demographic. Crucially, the ad watched nothing. It could not see whether you glanced at it, how you felt, or what you did afterwards. The information flowed one way — from advertiser to audience.

Surveillance capitalism reverses the gaze. The system watches you, continuously and individually, and the message is assembled for the specific person you are in the specific moment you are in — the exact mood, the exact hour, the exact vulnerability. And the loop closes: the system observes whether the nudge worked, learns from the result, and refines the next attempt. A billboard cannot learn from you. This does, at enormous scale, about you specifically. That shift — from a message broadcast at a crowd to a system that studies each individual and adapts — is the line between marketing as we knew it and surveillance as a business model.

It is worth being precise about one thing this article is not mainly about. Much of what surveillance capitalism funds is the engineered pull of the feed itself — the ranked, endless scroll built to keep you looking. That mechanism deserves its own treatment, and I have written about it separately in the attention economy. Here the focus is the layer beneath the feed: the extraction of behaviour, the manufacture of prediction, and the market where those predictions are sold. The feed is the shop window. This is the factory behind it.

Human experience as free raw material

The deepest part of Zuboff’s argument is not technical but moral, and it turns on a single move. Every extractive economy needs something it can treat as free — a resource it can claim, refine, and sell without paying the source. For the early industrialists it was land, forests, minerals: things declared available for the taking. Surveillance capitalism found a new frontier to claim in exactly this way. Its free raw material is human experience itself.

Your behaviour — your face, your voice, your movements, your moods, your relationships — was never anyone’s property to sell. It was simply your life. What surveillance capitalism did was declare that experience available for the taking, to be rendered into data, refined into prediction, and traded. The declaration was never put to us as a choice. It arrived as a fait accompli, buried in terms of service no one reads, defended after the fact by the argument that the data was just lying around unused.

Every extractive economy needs a resource it can treat as free. This one found human experience — your life, rendered into data and sold as a forecast.

This is where I find the pattern that runs through nearly everything I write about technology. It is a very old move wearing new clothes: something that was shared and un-owned gets captured and fenced off — enclosed — turned into private property, refined into a product, and rented back to the very people it was taken from. The common land of medieval England was enclosed with hedges and law. The common experience of daily life is enclosed with sensors and code. The machine is new each time; the move is always the same. I trace how that capture works across technologies in how technology gets captured, and the endpoint of the pattern — a world where a few platforms own the ground everyone else must operate on — in technofeudalism.

Ask the three questions I always come back to, and the shape is clear. Who takes? The platforms that extract the surplus. Who pays? You do — not in money, but in the raw material of your own behaviour, and in the loss of the right to a future the system has not already priced. Who fights back? So far, mostly researchers, regulators, and the occasional insider willing to say out loud what the incentives make inevitable.

What we learned when someone opened the box

For years this was an argument made from the outside, inferred from patents and public behaviour. Then insiders began to confirm it from within. The most consequential was the disclosure of a large cache of internal Facebook documents in 2021, which reporters came to call the Facebook Files. The broad picture they painted was of a company whose own research could see the tensions its systems created, and whose incentives nonetheless pulled toward maximising engagement — because engagement produces more behaviour to harvest.

I have written at length about that episode in Frances Haugen and the Facebook Files, and it is worth being careful about what it did and did not prove. It did not hand us a single smoking-gun equation. What it did was confirm, from inside, the structural claim at the heart of surveillance capitalism: that a business whose revenue depends on extracting behaviour has a built-in incentive to keep extracting, even when the people building it can see the costs. The specific figures have been debated. The shape of the incentive is not really in dispute.

Can you actually opt out?

The honest answer is: partly, and not by yourself. It would be dishonest to end an essay about a structural problem by telling you that the solution is more personal willpower. It is not. But there are things worth doing, and there is a reason to do them beyond your own privacy.

The individual layer is real and worth using:

• Prune the collection. Turn off ad personalisation, location history, and cross-app tracking in your device and account settings. On phones, the app-tracking controls genuinely cut off a lot of surplus at the source.
• Use tools that refuse. A tracker-blocking browser or extension, a private search engine, and a habit of denying apps permissions they do not need all shrink the exhaust you leave behind.
• Exercise the rights you have. Where data-protection law applies, you can ask companies what they hold on you, ask them to delete it, and object to certain uses. Even patchy laws give you levers most people never pull.
• Assume the surplus is the point. When a free service is unusually eager to reach your contacts, your location, or your microphone, treat that eagerness as information about the business model.

But notice the ceiling on all of this. No combination of settings lets one person step outside a market built on billions of others. The trackers follow you across sites you never chose. The predictions about “people like you” are built from everyone else’s data, so they describe you whether or not you personally opted in. This is a collective condition, and it can only really be undone collectively.

That is where the more serious levers sit — and where I would put my hope. Strong data-protection rules that make behavioural surplus expensive or illegal to hoard. Limits on what can be collected in the first place, rather than endless consent pop-ups after the fact. A legal recognition that some parts of human experience are simply not for sale, the way we long ago decided that people and votes are not for sale. None of these are utopian. They are the ordinary work of deciding, as a society, that a resource everyone assumed was free is not free after all.

Surveillance capitalism was not handed down by the laws of physics. It was a choice made by a handful of companies at a particular moment, when a new machine met an un-fenced frontier and no one had yet said no. The most useful thing to understand about it is that it can be un-chosen. The fence went up in our lifetime. There is no law of nature that says it has to stay.