In short
A dark pattern is an interface designed to get an outcome the user would not have chosen if the choice were presented plainly. They usually improve the immediate metric and damage retention, support load and reputation. Several categories are now explicitly regulated, including hidden subscription terms, obstructed cancellation, and consent interfaces that make declining harder than accepting.
The definition worth using
A dark pattern is an interface designed to produce an outcome the person would not have chosen had the choice been presented plainly.
That framing matters because it separates dark patterns from ordinary persuasion. Making a good offer prominent is persuasion. Making the decline option grey, small, and phrased confusingly is a dark pattern — the difference is whether you are helping someone decide or preventing them from deciding.
The common ones
Confirmshaming. The decline option worded to make the user feel foolish. "No thanks, I don't want to save money."
Roach motel. Easy to enter, hard to leave. Signup takes thirty seconds; cancellation requires finding a buried page, or a phone call during business hours.
Hidden costs. Fees appearing at the final step of checkout, after the person has invested effort.
Preselected extras. Insurance, warranties, subscriptions ticked by default.
Disguised ads. Advertising styled as content or as interface controls.
Fake urgency. Countdown timers on evergreen offers, "only 2 left" that never changes.
Consent manipulation. An Accept All button beside a Manage Preferences link leading to twelve toggles.
Nagging. Repeated requests for permissions the user declined.
The pattern across all of them: the effort is asymmetric. The outcome you want takes one click; the outcome the user wants takes five.
Why they backfire
They work, immediately. That is what makes them tempting and what makes the argument against them a long-term one.
The metric improves and the business does not. More subscriptions from people who did not intend to subscribe means more cancellations, more chargebacks, and more support contact. The signup number rises; revenue quality falls.
Support absorbs the cost. Every hidden fee and obstructed cancellation becomes a message someone answers. That cost lands in a different budget from the one that got credit for the conversion increase, which is exactly why these decisions survive internally.
Trust does not recover. Someone who feels tricked does not return, and does not weigh the incident against your other qualities. They tell other people, and reviews mentioning deceptive billing outlive whatever the pattern earned.
They compound negatively. A user tricked once treats everything afterwards with suspicion, including your genuine offers.
Several are now illegal
This has changed materially, and it is no longer only an ethics discussion.
Consent interfaces must make refusing as easy as accepting under European data protection enforcement — an Accept All button with no equivalent Reject All is a compliance problem, not a design preference.
Subscription rules in several jurisdictions require that cancellation be as easy as signup, and that recurring terms be disclosed clearly before purchase. Regulators have pursued specific companies over obstructed cancellation.
Consumer protection law has been applied to hidden fees and to interfaces that mislead about price.
If you are building for clients, this is worth raising as a risk rather than a philosophy — the exposure is real and it lands on them.
The honest alternatives usually perform
The useful finding is that transparent versions often perform comparably, and sometimes better.
Show total cost early. It reduces abandonment at the final step, which is where the expensive abandonment happens — someone who leaves at step one cost you nothing.
Make cancellation easy and say so. Visible reversibility lowers the perceived risk of committing, which increases signups. The number of people who cancel because it was easy is smaller than the number who joined because it was.
Offer extras clearly, unticked. Fewer people take them, and those who do actually want them — so refunds and complaints fall.
Give a genuine choice on consent. A clear accept and decline pair produces fewer accepts and far fewer complaints, and it is what the rules require anyway.
Use real urgency only. Actual stock levels, actual deadlines. Real scarcity is persuasive; invented scarcity is only persuasive until someone checks.
The test
Before shipping any pattern that improves a number, ask: would I be comfortable if the user could see exactly what we did and why?
If the answer is no, the pattern relies on the user not noticing, which means it works only until they do.
A second test, more practical: does the outcome we want take fewer clicks than the outcome the user wants? If so, that asymmetry is the design, and it is worth being honest that it was deliberate.
If a client is asking for something in this territory, book a call — there is usually a version that performs as well and does not create the liability.
Common questions
What is a dark pattern?
An interface designed to produce an outcome the person would not have chosen if the choice were presented plainly. The distinguishing feature is asymmetric effort: the outcome you want takes one click, and the outcome the user wants takes several.
Are dark patterns illegal?
Several categories now are. Consent interfaces must make refusing as easy as accepting under European data protection enforcement, subscription rules in several jurisdictions require cancellation to be as easy as signup, and consumer protection law has been applied to hidden fees and misleading price interfaces.
If dark patterns work, why avoid them?
Because they improve the immediate metric while degrading what it measures. Subscriptions from people who did not intend to subscribe produce cancellations, chargebacks and support contact — and that cost usually lands in a different budget from the one credited with the conversion increase, which is why the decisions survive internally.
Does making cancellation easy reduce revenue?
Usually not. Visible reversibility lowers the perceived risk of committing, so signups rise — and the number of people who cancel because it was easy is generally smaller than the number who joined because it was. It is also now a legal requirement in several jurisdictions.
How do I tell whether something is a dark pattern?
Ask whether you would be comfortable if the user could see exactly what you did and why. If not, the pattern depends on them not noticing. A more practical check: does the outcome you want take fewer clicks than the outcome they want?
