If you follow technology blogs, you have undoubtedly read scores of posts about best practices. Best practices identify tried and true strategies that will deliver positive results for most businesses within an industry, and they can be a great starting point when you are trying to revitalize your website. But the truth is that best practices can only take you so far. What works for a large global company might not work for your small business. Your target audience might respond differently than someone else’s. So how can you determine what will work best for you? That’s where A/B testing comes in.
What is A/B Testing?
A/B testing compares two different versions of a website page to determine which one will deliver better results. Site visitors are randomly shunted to one or the other of the page versions, and page metrics identify which version motivated more people to take the desired action. For example, if you want to improve the conversion rate for your landing page, you could create an A/B Test that compares variations in call to action wording to determine which version produces the most clicks.
Multivariate Vs. Single Variate Tests
You can use different techniques for your test, depending on what your goals for your website are. The two broad categories of testing are multivariate tests and single variate (A/B) tests:
- Multivariate tests compare pages in which multiple elements have been changed. They can be useful in testing a new page design or seeking to determine which elements have the greatest impact on user behavior. The primary drawback to multivariate testing is that it can be difficult to determine which specific change caused the results (either positive or negative) that you see. This can make it difficult to make decisions about future design changes.
- Single variate tests change just one page element at a time (for example, the wording of the heading OR the font size, but not both). These tests make it possible to fine tune your page and test specific elements to create better behavioral results. However, testing a new page design one element at a time can be a time consuming process.
The approach you choose will be determined by your goals. If you want to try out a whole new page design, multivariate testing may be the better approach. If you want to keep your current design, but gradually improve it for a better user experience, single variate testing may deliver better results.
What Should You Test?
When performing A/B tests, the most difficult part of the process may be deciding what to test in the first place. Here are a couple of tips to get you started:
- Use your site metrics to identify problem areas. If you see that a disproportionately high number of visitors leave your site at a particular point in the conversion process, begin your testing with that page.
- Test major elements first. Start with the most influential elements like the headings, call to action, and navigation buttons.
- Try many different changes to each element. If you are testing the call to action, for example, try changing the button placement, using a different button shape or color, altering the size of the button, making it 3D vs. 2D, or changing the wording. Just remember to make one change at a time to determine which has the greatest impact.
- Go back to that best practices list. Best practices lists can provide excellent ideas for testing. When someone recommends a certain strategy, test it!
Here’s one final tip for a successful testing strategy: don’t quit! Testing should be an ongoing process over time. As your audience changes, your home care website will change to accommodate their needs. Testing keeps you in control of the results, no matter how much user expectations alter.
Photo by adactio