A few months ago, I flew into New York City and hopped in a cab. When I got to my hotel, I swiped my credit card, and was a bit shocked at what I saw on the screen: the range of pre-selected tip amounts seemed really high! I started to wonder if the standards for tipping were different in NYC compared to Chicago… or worse… that I was simply a bad tipper!

Photo by Good Experience.

Photo by Good Experience.

After doing a bit of research and examining the interface, I realized that I was not a horrible tipper (whew!), and that the real problem lies in the use of UX “dark patterns”: deceptive elements that guide the user to behave in ways they otherwise would not. It turns out that other people here have had experiences with these dark patterns as well.

Below, we've called out the offending elements in the interface, and contrasted it with a different approach taken by a competitor:

Comparing and contrasting the "dark pattern" interface with a competitor interface I've seen in Chicago.

Comparing and contrasting the "dark pattern" interface with a competitor interface I've seen in Chicago.

  • By presenting a range of tip options skewed higher than the industry average of about 10%, the interface provides misleading information about the social expectation for tipping. Tourists may falsely conclude that tipping a cab driver between 20-30% is customary in New York City. Though the competitor's tip options are also skewed higher than the average, the difference is less extreme.

  • The exact tip amount is invisible until the user makes a selection. By not presenting the tip amount and final total alongside the pre-set tip buttons (as the competitor does), users are unable see exactly how much they are tipping on their bill until after they make a selection.

  • The interface hides the option of selecting a custom tip. Though the number keypad is displayed, no indication is given that it can be used to manually enter a tip. Users are likely to focus instead on the large buttons that display pre-set amounts. Contrast that with the competitor’s interface, which gives equal weight and screen real estate to the "other" option as it does to the pre-set amounts.

This interface reportedly has increased the average tip per ride to about 22% — a 12% difference — but it has done so by using dark patterns that exploit users. We certainly don't begrudge cab drivers a great tip, but we take issue when it comes at the expense of users — especially unsuspecting tourists or people who may be tired at the end of a long night! Rather, we recommend taking care to avoid design details that exploit users, and instead support them in making a conscious and informed decision to reward their taxi driver appropriately.