A/B Testing Inspiration to Kickstart Your Next Hypothesis [Infographic]

At this point, many of us are familiar with the term A/B testing, and how it has the ability to drastically increase website conversions, improve user experience, or encourage more website engagement.

Though using it, marketers are able to better understand user behavior so your business can begin tailoring your website to better resonate with your audience. Continuous iterations of these tests can help you learn what changes to avoid and what may be applied throughout other areas of your site.

But our websites are not one in the same. They fit into different categories and sell a variety of commodities, have different audiences, and maintain different interfaces. This means the tests you run for a retail company website could be fairly different from that of a marketing company's.

For example, if you are a retail company, you may find it more beneficial to test the visibility of reviews on products vs removing them. Or, if you’re a B2B company, you may want to test form lengths and questions to see whether or not your users like what you’re asking them.

To help understand what A/B tests you may benefit from the most, and how to do it, Optimizely has assembled an entire Testing Toolkit (complete with the infographic below) to help you and your team brainstorm data-driven tests. It even has a 'test idea framework' so you will learn step by step how to create a test and what to measure.

Crafting the Hypothesis

Prior to running your experiment, you need to create a prediction, or hypothesis, that states what you want to change, what you believe the desired outcome will be, and why. Your hypothesis should be a bold statement able to address possible solutions to a question you or others have. A proper hypothesis should be comprised of:

• “Problem Statement (or your reason for testing): Comprised of analytical data, user feedback, competitive review, etc.
• Proposed Solution (description of what to test): What variations you are testing and their differences, theories concerning why you believe this solution is correct.
• Qualitative Statement” (Measurable Outcome): What will be used to test the success of the experiment, how will your users be affected? Make sure to document your hypothesis so you have a record of your test to refer back to measure progress. This will also come in handy if you decide to share the results with your stakeholders or, decide to run similar tests again throughout the rest of your site.

Make sure to document your hypothesis so you have a record of your test to refer back to measure progress. This will also come in handy if you decide to share the results with your stakeholders or, decide to run similar tests again throughout the rest of your site.