When It Is Used Against You

A dark pattern is an interface deliberately arranged to produce choices the user would not otherwise make. Harry Brignull named the category in 2010 and catalogued the recurring forms, and the taxonomy has since been picked up by regulators: an unusual afterlife for a piece of design criticism.

The word deliberately carries the definition. A confusing interface is a failure. A dark pattern is a success at something other than what the user wanted.

The catalogue is short and recognisable. The roach motel: one-click subscription, cancellation requiring a phone call during office hours. It works on effort costs and the way friction compounds, since each additional step loses a proportion of people who genuinely intend to leave. Manufactured scarcity: countdown timers that reset on reload, stock counters unconnected to stock. Confirmshaming, where declining requires clicking something self-deprecating. Bundled consent, where agreeing to the necessary thing agrees to everything else. Pre-ticked boxes adding paid extras.

That last one attracts a defence worth examining, because it sounds reasonable. The box is visible, the argument runs, and anybody can untick it. The defence would be strong if defaults worked through concealment. They do not. They work through effort, implied recommendation and the status quo becoming the reference point, all of which operate perfectly well on a box in plain sight. Anybody choosing to pre-tick rather than leave blank has chosen the arrangement that produces more sales, and the defence describes the mechanism rather than excusing it.

Which brings this book to its uncomfortable conclusion. There is no technique in the dark pattern catalogue that does not also appear in the chapter before it. A default that enrols you in a pension and a default that enrols you in a subscription are the identical mechanism, working identically. Friction that stops you making an impulsive transfer to a scammer and friction that stops you cancelling are the same friction. The commitment device that helps you save and the small first request that starts a sales sequence are both foot-in-the-door.

There is no psychological distinction available. The only distinction is whose interest the arrangement serves, and that is an ethical question rather than a scientific one. Anybody hoping the research would sort the acceptable techniques from the unacceptable ones is going to be disappointed, because the research describes mechanisms and mechanisms do not have intentions.

This is also why the matter has moved from professional ethics into law. Most consumer protection assumes that disclosure solves the problem: tell people what is happening and they can act accordingly. These effects largely survive being known about: defaults still pull, friction still deters, urgency still works on people who can see the timer resetting. Disclosure is therefore not an adequate remedy, and several jurisdictions have concluded that the arrangements themselves have to be regulated. It is a rare case of a body of psychological findings becoming, directly, a matter of consumer law.