Think back to science class in high school. No, not the class where you had to dissect a frog! The one where you had to plan out an experiment, develop a hypothesis, and then you got to prove or disprove it.
Today, we’re going to apply that mindset very specifically to conversion rate optimization on your brand’s website.
What is A/B Testing?
A quick definition before we begin. An A/B test, also known as split testing, is an experiment in which we compare two variations of some element(s) on your website. You might think of them as the control (current version) and a variant or as variation A and variation B.
To test one against the other, the test itself is set to show each variation to website visitors so that some experience one and some experience the other. This test is typically set to 50% for each. After data is collected (via your testing platform or Google Analytics), the results will show either a winner or neutral outcome allowing you to update your website accordingly based on real data.
What to A/B Test for CRO
There are many elements that can make a difference for website conversions. Here are some options for your next A/B test:
CTA button
- Design (such as background color or font)
- Placement on page
- Size of CTA button
- CTA button copy
On-page copy
- Different headline
- Additional social proof
- Different copy or format of copy
Form
- Requiring fewer or different steps for form fill-ins
- Updated format of form
- Form placement
- Form headline
*Just remember – test one single element on any one page at a time. It can seem tempting to test several changes at once, but that doesn’t prove which change was the cause of the results.
Steps for running your A/B Test
1. Look for opportunity. Test where you think improvements can make a difference based on current data.
Is there a page or form that converts well for one service but not for another? Did your conversion rate fall recently? Does your Google Analytics show a lower-than-expected number of page views to your Contact Us page indicating that the CTA to get users there might not be working as it should? Whatever you find – digging into your analytics provides insights so you can make an impact.
2. Use an A/B testing tool.
We currently use Google Optimize, but as you may already know, it’s being sunsetted on Sep 30, 2023. Google has said it will be replaced with a tool within GA4, but word on the street is that they have been delayed in coming up with a solution that will meet security measures both in the US and Europe. Rest assured that if there is a lag in Google’s testing platform, there are paid options including Optimizely and VWO.
Note: You could certainly do a test without a platform, but using a platform is SO EASY. You can create your variations within the tool and it will track your visits and conversions and automatically end the test once it finds a statistically significant winner.
3. Make changes permanent.
Once your test ends, if your original version is the winner, there are no changes to make. If you find a new variant performed best, make that change permanent to your website. Additionally, consider if these test results are an indication of a change or changes that might work in a broader context or on similar pages or sections of the website.
4. Begin your next test!
Our Digital Growth program is all about building habits. So your next step is…Do another test! To continue to increase the Digital Value of your brand’s website, keep up the testing for continued, incremental improvements.
Conclusion
Every marketing team we work with faces the challenge of getting more online leads. We create websites using intentional design, architecture, and messaging for the purpose of converting visitors to contacts, but ongoing growth requires updates and changes based on learnings. A/B testing is a valuable tool for your marketing team to do just that. No matter how well your site is performing, testing will give you new insights for better conversion rates. It’s one test you’ll never fail and learn from every time.
