A friendly robot greets me on Facebook. He’s dressed like a doctor, stethoscope and all, here to do a security checkup. So for the next 5 to 10 seconds, I wait as he pokes and prods my account. “He’s really taking good care of me!” I think, when I start to wonder: Are Facebook’s servers really taking that long?
The short answer is no. Facebook actually slows down its interface to make users feel safe, a Facebook spokesperson confirmed in an email. “While our systems perform these checks at a much faster speed than people can actually see, it’s important that they understand what we do behind the scenes to protect their Facebook account,” the spokesperson wrote. “UX can be a powerful education tool and walking people through this process at a slower speed allows us to provide a better explanation and an opportunity for people to review and understand each step along the way.”
If half of Facebook’s billion users spend 5 seconds waiting on this check, that’s 694,444 hours, or 28,935 days of collective time lost. But Facebook isn’t alone. Websites and apps now operate on the magnitude of milliseconds. But such speed can make users skeptical or even confused, so companies are responding by building slower, more deliberate interfaces. Wells Fargo admitted to slowing down its app’s retinal scanners, because customers didn’t realize they worked otherwise, while various services on the web including travel sites, mortgage engines, and security checks are all making a conscious effort to slow down their omnipotent minds because our puny human brains expect things to take longer.
“Let’s say you sit down at a restaurant, you order your food, and it comes out one minute later. Is that a good thing?” asks Braden Kowitz, design partner at Google Ventures (which has more than 250 portfolio companies including Uber, Slack, and Nest). “You start to wonder, ‘What’s going on here? Is something wrong in the kitchen?'”
So companies introduce what Kowitz calls an “artificial waiting” pattern into their interfaces. These are status bars, maybe a few update messages, to construct a facade of slow, hard, thoughtful work, even though the computer is done calculating your query.
Kowitz says he has only used artificial waiting a few times in the design sprints he runs through Google Ventures, and only when instantaneous results weren’t working. In one case, he was working on a loan approval app. The back end was atypically fast. It could get someone a true lender-backed mortgage instantly. But when the app makers put that experience in the hands of consumers, people responded in disbelief. “When they saw it, they were like, ‘I’m pre-approved but not really approved,'” he says. So the designers added a progress bar that said it was checking credit, and suddenly, the same system seemed trustworthy.
Likewise, Steven Hoober, president at the consultancy 4ourth Mobile, was developing a service plan optimizer for a cell-phone carrier–a system that looked at a customer’s cell-phone usage and suggested the most bang-for-your-buck plan. “My guys built this tool–it took single digit milliseconds to get the results back. And it was giving [accurate] results, not just some plan we wanted to sell them,” Hoober says. “But when we tested with people, they assumed it was all marketing bullshit because it was instantaneous. They’d say, ‘This was obviously a canned result, I’m just gonna shop myself.'”
So Hoober introduced a waiting randomizer. The same code pulled the same results, but a timer delayed the display of those results anywhere from 8 to 30 seconds. The only visual feedback was a spinning indicator. Were focus group testers annoyed? Did they label the machine incompetent? Just the opposite. “People would sit there, asking, ‘I wonder what it’s doing?'” Hoober says. “Results would show up, and they’d lean in to the computer [assuming] it must have been really hard to get that information!” One hundred percent of people tested trusted the new tool, even though the only thing that changed significantly was their wait.
As subversive as this sort of design sounds, when used properly, consumers actually prefer the user experience of these white lies that take their time. In 2011, a team of Harvard researchers dubbed such phenomenon “the labor illusion.” They made up fake travel and dating sites to test how delays impacted a consumer’s reaction to the experiences. They found that people generally preferred the delayed results and considered them more valuable over instantaneous ones, provided that the interface included some level of “transparency”–feedback that explained the system was “comparing flights,” or whatever else it may have done in milliseconds.
There are limits to how long people will wait. The paper found that subjects were willing to wait up to 30 seconds on travel sites, but only 15 seconds on dating sites before waiting had a negative effect. “Although there may be a number of reasons for this difference (people may be more impatient to find a mate than a flight, for example), we suggest that one critical factor relates to consumers’ expectations,” the researchers write. Subjects expected that that a dating engine could find them a match faster than a travel engine could find them a trip. Was that expectation founded in some past experience of dating and travel sites? Possibly.
But that’s the point. Our expectations have supreme power to shape our experience of the world, including how we interact with our gadgets. “If we break customers’ expectations, the interface stops working,” says Kowitz. “That can happen if the Internet is moving too slowly. And in some cases, it can actually happen if the interface is moving too fast.”
In this sense, Kowitz believes that the industry can sometimes focus on removing points of friction–be that status bars, explanatory documentation, or extra screens–to a fault. If someone expects a process to take a long time, subverting their expectations with instant results isn’t necessarily a good thing, just like a steak set in front of you a minute after sitting down at a restaurant would raise the alarm.
“My guess is that over time, [our] expectations will catch up,” says Kowitz. “It’s one of those things where technology is getting a little ahead of what people expect, and that means we just have to meet them halfway.”