Follow us on Twitter
Grab the RSS feed
Page Color Has No Effect on Conversion
One of the questions we get a lot at Performable is: “Do light or dark web pages work better?”. This question usually comes after a design team has been discussing a change to their web site, and they’re considering their options.
There are several reasons why light pages seem like the right choice. First, the vast majority of web pages are white, and the history of publishing is mostly about printing on white or light paper, so white seems like the safe choice. White also lets you get a high contrast between type and background (black text on a white background), and in general high contrast is easier to read.
But a colored background also has advantages. For one thing, imbuing your background with your brand’s unique color can instantly communicate to viewers where they are. It is also nice to be different from all the rest of the Web. When you’re browsing white pages all day and land on a dark one, you take notice. And you can still achieve a high contrast between text and background with a dark web page.
So, given the reasons for using a light or dark page, we set up a test to find out which actually works better. We created two versions of the Performable home page, one with a dark green background and one with a white background. We kept the content the exact same on each page, the only thing we altered was the colors. (To do this test we used our own testing platform that we’re developing at Performable)
The result?
The results were surprising. The pages performed the same. There was no difference between the dark and light pages, even though they look entirely different. This was a surprise to us…we were leaning toward the white page because we thought it would appeal to a wider audience. (some people claim they just don’t like dark backgrounds).
We’ve written up the results in more detail over at ABtests: Performable.com light/dark home page test
A caveat: the most you can ever say about test results is that they are valid within the testing conditions under which they were run. Therefore, we can’t say that these results hold for all web sites and all audiences. It’s always necessary to test any hypotheses on your own audience before coming to more concrete conclusions. So, if you end up running a dark/light test of your own, we would love to hear about it.
Did you like that post? You'll love these.