The Standard · Article I

Deceptive practices.

A deceptive practice is a lie told by the interface rather than by the copy. Nothing on the page says anything untrue. The design does the deceiving.

Why this is its own Article. Article III covers untrue words — a claim you can't stand behind, a fraud given a respectable face. Article I covers something words can't reach. A countdown that resets on refresh doesn't state a single false sentence. It still deceives. That gap is what Article I closes, and it's why it's a red line.

The list

What we won't build.

Not exhaustive, and it won't stay still — these get reinvented. But if you recognise your own site here, that's the conversation.

The countdown that isn't real

A timer or a stock counter that resets on refresh, or that was never counting anything. Manufactured urgency to stop someone thinking.

Hard to leave

One click to subscribe, a phone call during business hours to cancel. Easy in, difficult out — deliberately.

The cost that appears last

Fees, surcharges, and handling revealed only at the final step, after the buyer has invested time and feels committed.

Already in the basket

A warranty, an insurance add-on, or a donation added to the cart on the customer's behalf, for them to notice and remove.

Already ticked

The marketing opt-in, the data-sharing consent, the recurring donation — pre-selected, so silence counts as yes.

Shaming the decline

“No thanks, I don't like saving money.” Wording the refusal so that choosing it feels foolish.

One button loud, one button quiet

Accept is large and coloured. Decline is grey text at the edge. Both choices exist; only one is offered.

The trick question

Double negatives in a consent box, so the customer cannot tell what ticking it means.

The trial that converts in silence

A free trial that becomes a paid subscription with no reminder, timed so it is missed.

Consent by exhaustion

Accept all is one click; refusing means six clicks through nested menus. Technically a choice.

Invented company

“Someone in Portland just bought this.” Fabricated activity popups, invented review counts, borrowed logos.

Friction on purpose

Extra steps added to the path the customer wants, and removed from the path you want.

Asking until they give in

The same popup on every visit, every page, until saying yes is cheaper than saying no again.

Advertising dressed as content

Sponsored material styled to look like an article, a search result, or part of the interface.

Where the line falls

Persuasion isn't deception.

Plenty of ordinary selling is fine and we'll happily build it. Real scarcity is fine — three left when three are left. A genuine deadline is fine. Highlighting the plan most people choose is fine. So is a good upsell, a well-made offer, and asking for the sale.

The difference is not how hard the design pushes. It's whether the design tells the truth about what it's doing.

The test. If you described to the customer, plainly and out loud, exactly what the interface is doing and why — would they still be fine with it? If yes, build it. If you'd rather they didn't know, that's the answer.

Not only a question of conscience

Regulators call these deception too.

This has moved from an ethics debate to an enforcement one. The US Federal Trade Commission has brought major actions over interface design — Epic Games paid $245 million in 2023 over designs that led players into unwanted charges. California's privacy law voids consent altogether when it was obtained this way, which means a consent banner built to wear people down may collect nothing it can lawfully rely on.

So declining this work isn't only us being particular. It's also us not building you something that becomes a liability.

Nothing here is legal advice. If you need that, ask a lawyer.