Conversion Killer Profile
The Design Problem
An unsightly creature. Uglow attacks website visitors on arrival, using ruinous color schemes, graphics and layouts to make people abandon ship. In extreme cases, he’s been known to induce seizures – not to mention lost conversions.
Aesthetics influences people’s perception of how well something works, like a website and the company behind it. Donald Norman studies the connection in Emotional Design:
“research on emotion and cognition has shown that attractive things really do work better”
In one study, for example, people reported that well-designed ATMs worked better, even though they were functionally identical to the less attractive ones in the test.
It comes as no surprise then that unappealing websites put people off. Here are some comments participants in my studies have said about websites they were testing:
It looks a little bit like it could have been done on a PowerPoint presentation slide.
The graphic design of the screen is really average to me. It looks about five, six, seven years old, which is a big issue I think.
I’d probably say it was a little dry, boring, doesn’t really pull you in.
It doesn’t say speed; maybe there’s kind of a lack of confidence there.
You can defeat Uglow by finding out – not guessing – what color schemes, layouts and graphic styles work best for your target visitors. Here are several ways to get the insights:
Use Existing Evidence
- Previous user research you’ve done: feedback, surveys, interviews, testing etc.
- Secondary research, like this
- Design patterns for ideas about what could work on your site
- Other websites your target market likes using – for inspiration
- Established competitor websites – for theories about what works
It is no bad thing to learn the craft by copying your elders and betters.
- David Ogilvy
Evernote’s site is a clean and simple design, for example:
Gather New Evidence
Do primary research:
- Capture feedback
- Carry out remote user testing
- Chat with visitors live
- Create a full survey or an onsite micro survey
- Recruit participants for interviews and usability testing
- Look at what people are saying on social networks, like Twitter
- Do guerrilla user research by getting anyone you can find – in a coffee shop say – to use your site and comment on the design. Any research is better than none!
First, do some follow up user research by asking people (ideally prospects) what they think of the intended color scheme, layout, branding etc.
Other peoples’ opinions may be interesting, but you really care what actual prospects think
Refine the prototype based on the feedback.
Next, turn your prototype into a final design, and test it against the existing design (the ‘control’) with real visitors using split-testing software like Google Content Experiments or Optimizely to see if it really is better for conversions.
Split testing software tells you when there’s a statistical winner (when the results aren’t just down to chance). If the new design turns out to generate significantly more conversions, from a stats perspective, replace the old design with the new one, making it your new control.