Data: The End of Web Design as an Art
April 01, 2016
I tested the color of a button on my company's website to see if changing its color would change the amount of signups that this particular website received. It did. This leaves me feeling a bit uncomfortable as a designer.
The issue is that I liked the color of the under-performing button better.
I could see myself not only testing the two colors that I tested, but also several thousand colors to eek out the best performing one. I could write a program to do this for me. I could write that program to work through other areas of the site to test those colors and sizes as well. And fonts, photos, and everything else. We could so fully optimize the site, that I'm not so sure that I'm even necessary any more. Perhaps I'm not.
I've always looked at web design as a form of art. It's what drew me to the profession in the first place, but if we're using data to inform every decision, it is no longer art.
I work for a fairly large company, where I expect that they are only interested in signups and results. I could look at it as the data confirms my artistic taste -- oh good, I can design a page that converts well. But the page I initially designed didn't convert as well as it could have. My first test -- changing that button color, confirmed that my design wasn't optimal, and the page that resulted from the testing was less appealing to me as a designer.
What if the optimal solution, as dictated by statistics and data looks like shit? Do I live with a website that I can't stand, but converts well? Do I place a website designed by data in my portfolio? I didn't design it?
Maybe I should just spend my time building out the software to build websites based purely on data and conversion rates, make my fortune, thereby giving me time to build sites that I want to build -- things that are beautful, but don't produce the optimal results
The Built Environment