I wanted to let you know about some recent tests we did for our client Jenn’s forthcoming cookbook Once Upon a Chef. One of the common discussions we have is whether or not to include the list of links to the retailers (Amazon, Barnes & Noble, IndieBound, etc) or to simply include a pre-order/buy button. Lots of authors include all the links and want us to too.
But, our design default in the past has been to include one clear button to purchase, with a pop-up which includes all the different retailer options (easily editable by our clients as more booksellers come on board).
For Jenn, we decided to test our assumption that one clear call-to-action to buy (with an additional click to get to the retailer) is better than a bunch of direct links.
Here are the three ads we tested using A/B/C split testing:
Which ad won?
The split test was statistically significant and Ad 3‘s click-through rate was 69% better than the other two ads. I was quietly pleased that my intuition about behavior was confirmed in the data.
Another click-through rate improvement
Moving the ad up from a low position in the sidebar to right below the site introduction tripled the click-through rate on desktop devices. (Note that it was important to segment out by desktop traffic only as the sidebar is below all the content on a mobile device, and a change to the sidebar is different on a mobile as the layout is different.)
- Run tests to check assumptions: Don’t blindly do what everyone else is doing. Test it out on your website, with your audience.
- Small changes can make big improvements: By running little tests, we can steadily improve click-through rates for key actions you want people to take.
- Reduce overwhelm: Keep your initial calls-to-action simple, the introduce additional options once they take action. (Getting people to click on buy first is the important thing, not where they’re buying from which can be overwhelming.)
Need help running A/B split tests?
You don’t need to buy any fancy software tools. Google Analytics, a bit of code we add, and a few minutes of my time to analyze the results are all you need. Let us know if you’d like to try experiments to improve a key action on your site.
Get actionable tips to grow your website
Thoughtful weekly insights (no hype!) on improving your website