The RunRex A/B Test Planning Guide
The RunRex A/B Test Planning Guide
Split testing or A/B testing is a process where a simultaneous experiment is done between two pages to see the page that performs better. The method validates any change to your website to improve the conversion rates. The process creates alternative pages for a site showing each of the pages to a predetermined group of visitors. In a classic A/B test, there are usually two versions of a page. The first version is the control, or existing design and the second version is the challenger page. A/B testing is part of scientific optimization techniques where the odds of your site visitors seeing the best performing version are increased using statistics. Some of the factors that lead to a poor A/B testing plan include ending your tests too early, changing more than one element at a time, failure to understand the significance and interpretation. The below RunRex guide is meant to point you in the right direction in coming up with the right A/B test.
Step 1: Measure The Performance Of Your Website
To improve your conversion rates, you should start with knowing how your website is doing and why its performance is where it is.
You can start by defining your site goals and objectives and set them up in Google analytics. You also need to define your KPIs and your target metrics. These indicators help you to make plans that are relevant to your business goals.
Once you have these measures in place, RunRex suggests you get feedback from your visitors. You can set up surveys, segmentation or use Google analytics to get qualitative data and answers why your visitors visit the site and what they do not like about it. Check for the most common trend in your customer’s feedback to know what might be the problem.
Step 2: Prioritize Your Testing Opportunities
You now have your goals and metrics in place; the next step is to choose what to test. Start from testing opportunities that can give the biggest gains. You can also prioritize what to test using data. For instance, Google analytics data might show you some of your site pages that get visitors than you would expect. You should also consider prioritizing pages that have a potential for improvements such as problematic pages or those with high bounce rates. Additionally, prioritize tests that are based on cost and value. Look out to prioritize pages that are:
- Most visited pages
- Top landing or entry pages
- Pages with expensive visits
Consider the changes you are going to make and consider if the change will have an impact, if the change will be noticed under 5 seconds, if it adds or removes anything and if the test will run on high traffic pages.
Step 3: Testing
The reason for the above step is to get an understanding of what is happening to the website and look for areas where A/B testing can provide results. Coming up with a test involves:
- Come up with a hypothesis
Your hypothesis or theory is the foundation for everything else you do in an A/B test. Thus, it should be measurable, clear and concise. Think about what the element you would want to change and what actual success would look like to you and why a problem occurs on your site. Think and research on what you would want to test and if there is evidence for your hypothesis. The hypothesis should be measurable, have a goal and give you essential information about the customers.
- Do not ignore the statistical confidence of your A/B results. The statistical confidence of the test should also be 95% as the measure of confidence so that you know your results are valid.
- You should also test for revenue. Thus, it is important to track the correct indicators to make sure that an increase in conversion rates does not result in a decrease in revenue. Some of the elements to test include:
- Call to actions
- Images product pricing
- Amount of content on the page
Step 4: Learn From The Results
You should have a plan for constant improvement. Sometimes not all split tests are successful, but a split test can give you a baseline to improve on. After some spilt tests, you will start to see improvements in your website.
Some of the other considerations that RunRex points out that you need to remember as you do your A/B testing include:
- Be aware of the local maximum
- Remember to adhere to the SEO considerations such as:
- Do not run your experiments longer than needed
- Instead of using no index Meta tag on your variation page, use rel=canonbical
- Use 302 instead of 301
- Do not use A/A testing
The right A/B testing plan allows you to increase your site revenue. Additionally, the plan gives you additional insights about your customers such that you are aware of their preferences instead of guessing what your customer might want. Further, the right A/B testing plan indicates the most important elements of your website. The A/B testing plan gives you evidence of what works in your marketing plans. You should continuously test the hypothesis to get better conversion rates.