Anyone who has even breathed in the vicinity of a website redesign knows that there is no more perfect time for every single site user to say exactly what they think about the terrible, horrible changes that never should have been made in the first place. Now consider Facebook, with its 1.32 billion subscribers, most of whom feel quite proprietary about the site. Who wants to mess with a billion avid users?
Margaret Gould Stewart does. As Facebook’s director of product design, it’s her job to make sure the site’s latest layout changes don’t tilt the love-hate Facebook equation too far toward “hate.” The Silicon Valley veteran, who led user experience teams at YouTube and Google before joining Mark Zuckerberg’s crew, says that designing for the masses brings up particular challenges: Any pixel tweak could rankle half the population of the internet. Here’s how she does it:
1. Sweat the small stuff. Obsessively. How hard is it to redesign something as small as Facebook’s “Like” button? Very, Stewart says. The design team assigned to the job had strict constraints: The button had to conform to specific height and width parameters, be understandable in different languages, and eschew fancy gradients and borders that would degrade in old Web browsers. The lead designer estimated that he spent more than 280 hours on coming up with a solution; the entire project spanned six months. That’s actually not that long when you consider that the tiny button is seen on average 22 billion times a day across 7.5 million websites.
2. Beware the false positive. When analyzing the resulting user data, the company holds off on declaring a new feature a success until it can dig beyond what might seem to be a positive response. “When you launch that interface with a new button that was never there before, you’re going to get a bunch of clicks just because it’s new and in people’s field of vision,” Stewart says. “It doesn’t necessarily indicate what the long-term response rate is going to be.” Wait until the novelty effect wears off, she cautions, and then take a closer look at whether the change enhanced or compromised the user experience.
3. No surprises. If Facebook decides to do a broad rollout of an update, it’s careful not to spring it on users without warning. Instead, it explains the rationale behind the change and gradually migrates users to the new format. “Helping people to understand that change is coming is great because they can emotionally prepare themselves,” Stewart says.
4. Remember how different users can be. In Facebook’s case, its product must be as accessible to a San Franciscan with an iPhone as to a person in New Delhi with a low-cost Motorola. “Designing for low-end cell phones is not glamorous design work, but if you want to design for the whole world, you have to design for where people are, and not where you are.”
5. You can’t please everybody. But all the testing and forewarnings can’t prevent some users from getting ticked off, for one simple reason: Nobody really likes change. “People become very conditioned to using a particular design, even if it’s not well-designed,” Stewart says. They adjust to the quirks of poorly designed software, for instance, and develop muscle memory around them. Just imagine if a command you use on an enterprise program every day suddenly appeared on a different dropdown menu—cue teeth gnashing. “When you’re receiving feedback, you need to figure out ways to separate what is change aversion—the ‘Who moved my cheese?’ response—and which are the substantive bits that say, ‘This is truly tactically not working for me anymore.’” That’s not easy to parse, and Stewart recommends waiting a few weeks for users to acclimate before recoiling in humiliation. If most of the feedback is just a negative reaction to change, that’s a win.