"Make it easy for the customer to say yes." You hear it at every conversion talk, every UX audit, every checkout redesign. Remove friction. Grow the yes rate. It sounds right. Hopper followed the advice to its logical end, and the FTC settled with them for $35 million.

Here is what the advice looked like in practice. A woman books a flight on Hopper and sees two things: a total price and a green Swipe to Book button. Below the scroll line, out of sight, two fees sit pre-checked and already added to her bill. A Tip fee. A VIP Support charge. Both checked before she touched a thing. Both billed the moment she swiped. She never chose them. She paid for both.

There is a name for this in the literature. And it did not start with Hopper.

Eric Johnson and Daniel Goldstein published a paper in Science in 2003 called "Do Defaults Save Lives?" They studied organ donation rates across Europe and found a split that no attitude survey could explain.

In Germany, about 12% of adults are registered organ donors. In Austria, next door, the rate is near total. Same language. Same border. Same religious roots. When you ask people in both countries how they feel about donation, the answers line up almost point for point.

The gap comes from one line on a form. Germany asks you to opt in. Austria signs you up and asks you to opt out. Same people. Same stated beliefs. One checkbox producing a wildly different outcome. Johnson and Goldstein popularized this as the default effect. The mechanism does not persuade. It selects for inertia. The person does not decide. The form decides for them, and they let it stand.

Hopper's product team took that same force and built it into a booking flow. The fees were on the screen, technically. But they sat below the scroll line, past the button the user had already been told to press. The design did not hide them in the legal sense. It hid them in the human sense. You would have to go looking for the charge you were already paying.

Then Hopper tested it. Their own A/B data told the full story. When a fee was shown openly and left unselected, 15% of users chose to pay. When the same fee was hidden below the fold and pre-selected, 75% paid. Five times the conversion. Same product. Same price. The only change was which box was already checked.

That is not a conversion gain. That is the default effect doing the selling.

Over three years, Hopper collected roughly $85.6 million in Tip and VIP Support fees with this checkout design. The revenue was real. So was the internal pushback. One Hopper employee put it in writing: "To me, the problem here is that we're tricking users."

The FTC agreed. Hopper settled for $35 million in early July. The company will be barred from using hidden, pre-selected fees going forward.

And Hopper is not an outlier. An ICPEN review of 642 subscription sites found that 76% used at least one dark pattern in their sign-up or checkout flow. The mechanism is standard practice across the field. The settlement did not reveal a new tactic. It gave an old one a federal price tag.

Most people treat this as a moral failing by one company. It is a design flaw baked into the advice itself.

The platitude does not mark where the line sits. It does not separate removing friction on the path to a choice from removing the choice itself. One builds trust. The other borrows from it. That is its failure point. The advice is not wrong in spirit. It is broken at the level of design, and the broken version pays better in the short run. That is why it spread.

Make the choice easy. Make sure there is still a choice. Once that is clear, three moves follow.

Open every checkout, sign-up, or payment flow you control. Find what is pre-selected before the user touches anything. For each checked box, ask one question: would a reasonable person pick this on their own if the box were blank? If the answer is no, blank it. Most operators skip this move because the lost revenue feels immediate and the risk feels far off.

Load your checkout on a mobile screen. Note what sits above the fold, the part the buyer sees before scrolling. If any fee, add-on, or charge lives below the action button, move it above. The buyer should see the full cost before seeing the button that locks it in. If you would not want to find a hidden line item on your own phone bill, do not put one on theirs.

Pick one add-on or optional fee in your flow. Turn off the pre-selection. Leave the product shown, described, and one tap away. Track the take rate for 30 days. The number that comes back is the real demand. The gap between that number and your current rate is the share of revenue the default, not the product, was earning.

At the end of 30 days, ask three things.

→ What held its take rate without a pre-selected default?
→ What looked like revenue but was really the checkbox doing all the work?
→ What friction showed up more than once when buyers had to make a real choice?

That is the difference between advice that sounds right and a system that proves itself.

Running this for 30 days does something the old advice never did.

You see which products carry real demand and which ones only sold because a box was checked. You see where your checkout was building trust and where it was borrowing from it. You see whether your conversion numbers reflect a real yes or just the weight of inertia.

The default effect runs whether you name it or not. That woman who booked a flight never chose the charges on her bill. Hopper's checkout earned $85.6 million in three years on fees like hers, and cost $35 million to settle. Borrowed revenue always comes due.