3 Aug 2015
I was recently giving guidance to a new designer about a website’s navigational architecture. He’d run into both technical and design constraints, and I convinced him to abandon the complex structure he was trying to use for a simpler layout.
“I know it’s better,” he said, “but I’ve spent three days trying to get this to work. It’s a shame to abandon it.”
It’s hard to walk away from a project. Sometimes we don’t get a choice; projects get cancelled, funding get shifted, staffing changes, assignments get juggled, technologies change. Of the first five projects I was assigned as an Information Architect, only one made it to production. Other times the project was one of my own making, a complex mess of decisions that added up to a problem in the design or the development. Wishful thinking and stubbornness are a great combination if you want a result that doesn’t actually work for anyone.
Continuing to work on the wrong solution just because we’ve spent so much time on it is an example of a sunk cost fallacy, the phenomenon where we justify increasing investment in a decision despite evidence suggesting that the benefits aren’t worth continuing. Letting go of a commitment we’ve already made is extremely hard, even when it’s what we have to do.
I can remember making the same complaint about a cancelled project when I was a new designer that my acquaintance was making now. “It’s a shame to abandon it.”
“Yeah, but you learned something, and you got paid,” my mentor replied. “You did what you were asked to do, and you’ll be a better designer for the next project. There will always be another project.”
Eight years and dozens of projects later, it’s still true. A day’s work learning something - even if it’s learning what not to do - is money well earned. The privilege of getting paid to learn to be better designers, even when that wasn’t the intention, is a pretty good deal.
Today, I learned something and I got paid. Whether that means the project is cancelled or not, whether it ships or not, the time wasn’t wasted.