1. Using Oracle CX Commerce Experiments
  2. Understand Experiments
  3. Plan experiments

Plan experiments

You can use Oracle Commerce Experiments to enable many different kinds of tests. A vital part of performing these tests is to ensure you plan the experiment before you actually run it.

You should keep the following points in mind when carrying out A/B testing on your store:.

  1. Include A/B testing into your everyday workflow process, especially when planning new promotions and landing pages, and optimizing your site and marketing channels, such as email. A/B testing gives you the opportunity to learn more about what does and does not work with every campaign.
  2. In addition to analyzing individual A/B tests, make sure to do a regular review of A/B tests overall, so that you can identify patterns in the results. Over time, you will gain a better understanding of the kinds of tests that work best for different circumstances, and be able to apply learnings from one channel to another.
  3. As A/B testing is unbiased and provides data-driven results, it is a great way of resolving internal disputes. Utilize it as a way of aligning multiple stakeholders around common goals and data points to drive improved performance and priorities.
  4. A/B testing is iterative and will sometimes require multiple tests to be taken in order to get useful results. Therefore, you may need to rerun tests as your business, customers, and market evolve.
  5. You should run one experiment at a time to avoid uncertainty around identifying which changes led to a change in results.
  6. You should use A/B testing in tandem with other tools and methods, for example, web analytics, heat map analysis, and user testings to better understand the ‘why’ along with the ‘how’.

In order to run an effective experiment, it is important that you choose the right success goals and have a hypothesis in advance. For instance, you may wish to run an experiment to test what the impact of using a video on the product details page has on clicks/add to cart, or on the conversion rate. Or you may run an experiment to test whether the placement of a promotional message on the cart page increases the average order value, revenue per visitor, or conversion rate.