Split Testing – Knowing is Half the Battle, Part 2
This is part two in a series of four posts on understanding the realities and challenges of split testing to improve your website’s conversion rate. If you missed part one of the series, you can read it here.
Once you’ve determined that your site gets enough traffic to support a reasonably short testing cycle (if it’s going to take six months, you may want to reevaluate your test), you’re ready to flesh out your testing ideas and go for it. A word to the wise, however:
Coming up with ideas is the easy part. Coming up with good ideas is a little bit harder.
When you catch the bug and get excited about split testing on your website or external landing pages, it will be very easy to jot down a lot of ideas for things you’d like to test.
As you’re compiling these ideas, make sure you’re focused on improving the user experience rather than hoping that “different” will magically equate to “better.” Lance Loveday, conversion expert and co-author of “Web Design for ROI” calls this “testing among good options.” Unfocused testing wastes valuable time and opportunity, so make your tests count!
One of the most common examples of unfocused testing is changing colors – button color, logo color, general Web page color schemes, etc. Once in a while, a color test may be appropriate and effective, but this is the exception rather than the rule.
A good example of an exception in this case would be if the color of a button impedes the ability of a user to easily recognize it on the page or read the button’s call to action. I recently discovered some yellow buttons with white text that were somewhat difficult to read without squinting. In this case, it would be wise to test a darker button color or text color to see if darker buttons would produce a significant lift in click-throughs.
However, it’s a good bet that this sort of test would not be a good idea if you’re thinking, “I’m tired of this light blue. Let’s spice things up and try a deep red instead.” Changing colors purely on the basis of personal aesthetic preference is not likely to make an impact on your website conversion rate.
The Takeaway
You’re more likely to have success testing intelligent changes to elements of your website that have a significant impact on the user experience: your headlines and messaging, page layouts, calls to action, and other elements that impact a visitor’s perception of your benefits and ability to fill needs and add value.
Tags: Conversion Optimization, cro, split testing
True. There has to be a hypothesis behind the test. For example, we added a try before sign-up bar on our homepage of http://www.reedge.com with the idea “if people can try part of the software before sign-up they are more likely to start using the testing tool after sign-up”.
So with a clear goal in mind we re testing this idea for 80% of the site visitors and a 20% control group.
Thoroughly enjoyed this post very clear down to earth content. Especially being aware of color changes based on user experience rather than personal choice. Let the market decide rather than letting the ego have a word in edgeways;-0