Preview Mode Links will not work in preview mode
  1. All Episodes
  2. Everything You Know About Conversion is Wrong with Chris Dayley, Ep. 56

Everything You Know About Conversion is Wrong with Chris Dayley, Ep. 56

Jul 26, 2018

Today’s episode of Just the Tips finds me recovering from an accidental 22-mile walk, but that doesn’t stop me from getting excited about our guest, Chris Dayley. Chris is a mastermind at conversion, having done more than 1,000 split tests in his time, and applying psychology to marketing to really help businesses dig into what makes customers convert. On today’s show we talk about the assumptions people make about websites, how most of them are wrong, and how to run a proper test to grow those conversions. This is a hugely valuable episode of Just the Tips that you won’t want to miss.

Traffic rises, but conversions stay stagnant

Chris’s story of how he got into conversion is a really great lesson for anyone out there trying to parse what analytics mean the most to their business. He was in-house at a company and focused on organic traffic. And in fact, his team tripled the company’s organic traffic in just six months, which is a phenomenal climb. But as the traffic rose, the conversions remained stagnant, and no one could tell Chris why that was. And as Chris says, if you’re getting lots of organic traffic but few conversions, there are basically two problems: Either you’re showing up for an irrelevant search term, or there’s a problem with the website. And so Chris created a split-test that looked worse than the current website, but worked better. Why? That question has fueled Chris’s career, and provides the launchpad for a fascinating episode of Just the Tips.

What works best doesn’t always look best

What Chris found so fascinating about his test is that the page he created looked, aesthetically, way worse than the current site. So much so that the designer said he would never put it on the site. But still, it converted better. We often make the mistake that the most important thing about a website is that it looks good. But really, you want something that works well. Chris tells a really interesting story about something he calls “an existence test,” which you’ll want to hear about. It’s really savvy stuff you’ll hear on this week’s Just the Tips.

Pick and choose

Chris told us about how he helped a client sell $25,000 more in one week through one of his split-tests. How? He says that businesses tend to make a lot of dangerous assumptions about their customers and websites. And sometimes there is data behind those assumptions, but the data doesn’t tell the whole story, or we read into the data the story we think is correct. The other assumption people make about websites is that they think they need to put everything on there: Dump everything onto the site so that there’s something for everyone. But as Chris says, his split-tests have shown that old theory is wrong.

Run the test

One thing that Chris said that really stuck with me is that you have to build the site that your audience wants. They need to tell you (through their behavior) what they want in your site. So even though simplicity is often good for a site, sometimes your audience may not want the site to be so simple. It’s all about throwing out the assumptions that we make without doing the hard work of split-testing. As Chris says, even customers may not know what they want. You can’t ask people what they want on the website. You have to run the test. Listen to the episode!

Outline of This Episode

  • [5:15] How Chris got started studying conversion
  • [9:26] What looks best doesn’t always convert best
  • [12:45] An existence test
  • [17:41] The psychology of conversion
  • [24:45] Why our assumptions are so bad
  • [29:43] What’s a good conversion rate?
  • [32:50] Chris’s toolkit

Musicfor “Just The Tips” is titled, “Happy Happy Game Show” by Kevin MacLeod (http://incompetech.com) Licensed under Creative Commons: By Attribution 3.0 License

Resources Mentioned

Connect With James and Dean

James P. Friel:

Dean Holland: