Thanks to deeply-ingrained core training, I generally start design projects of any kind with the broadest of objectives (what’s the bird’s eye view?) and drill down to a fairly granular level before putting time into real design work. I’m always amused when clients say, “Just create a design and we’ll come up with content to fit.” ERROR. (bzzzt, crack) DOES NOT COMPUTE.

There are occasional exceptions, though, and I believe that those exceptions are often critical to real breakthroughs. A good designer needs to keep a small percentage of their mind open to whimsical, random, out-of-the-box ideas in order to avoid an end result that’s bereft of any spark. Most design needs to be, first and foremost, functional and efficient. But if you want people to enjoy using it (put another way: if you want them to keep using it), it also needs to be fun. Purely data-based results aren’t generally the life of the party.

Ultimately, though, the success of a design has to depend on it achieving user goals, whether those goals include fun, efficiency, insight, or all three. It’s easy for designers at any level to become married to an idea and then feel like the data disproving that idea is flawed. If the designer in question is Steve Jobs back from the dead, okay, maybe he’s right. Otherwise, we all need to have the humility to back up, realign our understanding and expectations, then move forward with a plan that the data does support (still being open to design magic!). 

Once a strong, data-backed plan emerges, it’s time to start playing with options (lots of them) and sweating the details (ALL of them). Fonts, colors, effects, movements… the more potential bases you can cover, the better your chances of landing on the golden chalice of design: iconic. Such a status doesn’t necessarily even mean that end users notice it, just that they love the hell out of it.