Learning Tree just published my article on why charts (and other data-intensive components) are usually a sign of lazy UI design.
I was chatting to Sean Brady of Prism Decision Systems last week. He was using the Modern UI version of Skype on his new laptop and bemoaning the fact that software can be too minimalist.
There’s a recent trend for software to have minimal options. Design decisions are made for the users by the developers. iA Writer, a word processing application, typifies the genre. It famously doesn’t allow users to change font type, size or color—you just sit down and get on with it. Contrast this with Microsoft Word which has more options than Macy’s lipstick counter.
Software that minimizes the options available to the user is called “opinionated”. Options that, in the opinion of the developers, don’t add much value to the user experience, should be removed. In fact, some options can actually degrade the user experience—many a Powerpoint author would have done well to stick with the default options.
Does the average Word user really need dozens of fonts? Can any mortal use more that 10% of the features of Adobe Photoshop? What value is there in changing how your tabs look in Internet Explorer?
Research has show that too much choice can be a bad thing. Choosers become overwhelmed by choice paralysis. Don’t try and pretend that you’ve not whiled away many a corporate hour endlessly tweaking a presentation slide, unsure of what looks best.
But, how can more options be a bad thing? What if we hide them away behind a single “Settings” button nestled in the corner? Then the majority of users can have simplicity while the power-users tweak the shading on the scrollbars.
One of the problems is that these options add complexity. They bloat the software. Each addition is a potential source of bugs and, as the application is extended—or even redeveloped—these options begin to weigh down progress. There’s a finite amount of time available for design and implementation. If some of this is being squandered on options of questionable utility, it’s taking time from core innovation.
There also the problem that the developers themselves lose their way. Options are generally added to placate a user. Rather than spend time thinking about the value of a request, and explaining to a passionate user why it won’t be implemented, developers often take the cowardly step of adding it in as an option. Effectively this is saying, “I’m not sure if it’s valuable, but we’ll just give you the option to do it as that’s less confrontational.”
As a developer, if you do that enough, you begin to lose the vision you had for your product. It stops being a handcrafted work of art and becomes a consensus-ridden community product. Believe me. You really don’t want developers to lose their passion for your favorite app.
The danger is to think of opinionated software as mininalist. It’s focused. Just complicated enough to do the job you need done—and no less. I often remind my clients that what they really want is a big green button that they press and out pops the answer. The design question is how close we can get to that. They rarely need an elaborate chart—they need conclusions drawn from that chart. They rarely need a sortable, pageable grid—they need a “suggest-as-I-type” search box that pinpoints the records they are hunting for. Opinionated software melts into the background while you get on with the task.
One of the benefits of the explosion in low-cost, simple apps is that you can find one that the perfect fit for you. No longer do we all have to use the same behemoth applications and tweak their edges to find some personality.
And, if all else fails, in this age of the citizen programmer, if you can’t find the app you want, build you own.