Notes From Getting To Consistency: Don’t Make Your Users Think
SXSW, Saturday morning
Panelists: Paul Schreiber from Apple; Jennifer Fraser, Corel ; Alex Graveley, VMWare and Steve Johnson, Adobe.
Products don’t have to look the same, but they should perform the same. When you do something in one program, the same sequence of keystrokes or clicks should do the same thing in another program. Example of shortcut keys, which frequently don’t work the same from program to program.
Example of the USPS automated postal counter. It makes you answer yes/no in awkward sequence. “Is there any question that the package will fit? Will your package fit?” First question sets you up to answer “no,” but second requires you to answer “yes.”
Consistency doesn’t mean staying consistent with your entire legacy base. Apple was smart in ditching the floppy drive and just moving on. You shouldn’t let the needs of a very small number of users constrain you from innovating.
There are costs to inconsistency. Tech support costs are higher. You may actually alienate customers if they believe that you’re ignoreing their platform or designing an inferior product for it. You could incur costs to reverse-engineer consistency later.
Electronic Arts hasn’t changed the UI for Madden Football since version 1. What’s the customer’s goal and what can we do to help them achieve that goal as quickly as possible? If you can do that using consistency that users expect, then that’s great. But if you have to break a sequence to achieve a goal for the user, that’s what you have to do.
Go out and watch customers. If what they’re doing to achieve something doesn’t make sense, redesign it.
Adobe saw wide-screen monitors coming into widespread use and so provided a way to easily reconfigure the UI for different aspect ratios.
But they won’t always tell you what they want. Malcolm Gladwell was recently talking about spaghetti sauce. He said that spaghetti sauce makers used to think there was one ideal kind of spaghetti sauce. But it turned out there were different groups of tastes that people wouldn’t admit to. They liked chunky spaghetti sauce but didn’t think to say that. Prego figured this out and made hundreds of millions of dollars.
Who are your users going to be? You don’t need to be consistent between interfaces for a fourth grader and a legal secretary. Likewise, when you introduce something new, do you do it for your new customers or existing customers? Different expectations if you’re introducing something that helps people get started with the product. In that case, you don’t need to consider consistency with previous versions.
Cross-platform considerations. How much do you make it look like your product on another platform and how much like the other platform’s conventions. There’s “OK-Cancel on Windows, “Cancel-OK” on the Mac. You need to conform to these conventions. VMWare is creating its first Mac product and has had to revisit its whole approach to interface design to make the experience consistent for Mac users.
Features are the F-word. This was a point Steve Johnson made. Engineers fall in love with new features, but features can disrupt the user’s experience. Are you introducing features for the user or because you think it’s cool?