Click here to preview the new Fast Company

Want to try out the new

If you’d like to return to the previous design, click the yellow button on the lower left corner.

How Do You Know When A Design Is Really Done?

Should you move fast and break things, or methodically research and develop? Mobile design expert Luke Wroblewski weighs in.

David Fincher did 99 takes of the first scene in The Social Network. James Dyson made 5,127 vacuum-cleaner prototypes. Meanwhile, Paula Scher sketched out what became the Citi logo on a napkin on the first try. How do you know when a design is done—neither over-analyzed nor too facile? It's a question that doesn't just apply to product designers, as our latest advice-seeking reader knows:

"When you're making something (whether it's a family Christmas card or a consumer product), how do you know when you've done enough testing? Or too much? Put another way: How do you balance a thirst for data with your gut instinct?"

The tension between form and function, taste and testing, is unavoidable in design. Skew too far in one direction and you get Google splitting hairs over 41 shades of blue; too far in the other, and you get the Mac Cube. Finding the middle way can be a fraught process.

But it doesn't have to be, says mobile designer Luke Wroblewski. No stranger to rigorous testing, Wroblewski has to make sure his designs work on dozens of different devices. But he's also unafraid to commit to idiosyncratic choices, like the odd split-screen interface for the desktop version of his polling app Polar.

What's his secret? Data and intuition, he says, should "complement each other, not compete." The way to put that maxim into action is to forget about aiming for some objective bulls-eye of perfection—all due apologies to Steve Jobs—and treat the act of creation as a continual learning experience that empirical research and gut instinct both inform.

Citibank Logo via Flickr user BruceG1001

"The more tangible, real-world data you drink, the more you fill your gut with the right kind of instincts," Wroblewski says. "The way I’ve personally managed this balance is by thinking of launches as part of the design process. What you ship is a timestamp. At this point and time, I have applied what I know into the design. After launching I’ll learn more and apply that to create another timestamp. Think of it as your Christmas card getting more awesome each year. It’s inevitable that you’ll look back on some of your earlier attempts and think, 'Wow, that was bad.' But it was right based on what you knew at the time."

Note that Wroblewski is not advocating that you avoid uncomfortable feedback, nor that you avoid making decisions until you've asked everyone you know. He's advocating that you discard certainty as a prerequisite for action. Certainty means there's nothing left to learn, no ambiguity or risk to accept. Are you really trying to make the best thing you can—using your own gut instincts and observable facts to guide you—or are you simply trying to avoid uncertainty?

Gut feelings aren't infallible; neither is data. Stop pitting them against each other in a zero-sum truth tournament. Instead, make them friends that teach each other. You want certainty? Wait for death and taxes. You want to make great things? Intuition and information can both point the way, but you've still got to put one foot in front of the other.

[Image: Fiber optics testing via Wikipedia]

Add New Comment


  • Viktor Kis

    The ultimate quote is in the article itself, "discard certainty as a prerequisite for action". Great line.

  • This is typically the role of a manager - knowing when to "shoot the engineer/designer" - and move on with things. In terms of product design where safety and reliability come into play, clearly those two issues have to be addressed to the best of our ability, given the data we have. Beyond that however we have some flexibility for improving and refining the product after it ships. The basic question that needs to be answered is "is the product (design) good enough?" and by that I mean "does it solve the basic problem/need?"

    I will end with a quote that I like: "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery French writer (1900 - 1944