Overview: A/B testing

Optimizing the digital experience on your site and making marketing decisions, based on solid data are key catchphrases employed when talking about converting visitors on your site to something more than visitors. You may want your visitors to submit a form on a landing page of a campaign, so that you acquire contacts information. Or, for your campaign it may important that visitors navigate to a specific page to download your pricelist, and so on.

So, how do you know whether your landing page or new call-to-action button design are going to do the job and convert contacts? To shift your marketing strategy efforts from assumption to knowledge, you can run A/B tests to experiment with variations of the original page to measure which variation performs better in terms of making visitors complete the desired goal. Page variations are displayed to visitors at random at traffic distribution that you set for a specific duration. Next, you explore performance results and statistical analysis for the effect of the change, based on which you decide which variation is your winner that optimizes the page to the desired outcome. You can experiment with almost any part of your page – forms, layout, position of page elements, content elements, call to action, images, and so on. Thus, with A/B tests you validate and refine your marketing strategy.

A/B testing is a part of a wider ecosystem of optimizing your website and increasing conversion rates. For example, website analytics may help you identify problems with visitors' experience on your website. The case may be that bounce rates are high, or not enough visitors go your promotion page and leave their contacts. In combination with your business objectives and reasonable metrics, A/B testing is a key method to optimize conversion rates and make data-driven marketing decisions.

The general A/B testing workflow is as follows.

1. Develop a hypothesis for your test.
   The hypothesis basically states the reason you think a problem occurs. That is, what you are testing and to what forecasted result. Say the problem you identified is that you do not have enough leads to boost sales on your site. Your hypothesis may be that visitors cannot easily subscribe to your newsletter, so making subscription options more prominent by placement and size will enable visitors to complete the process more quickly and easily.
2. Identify the goal you want your visitors to complete.
   For example, on your DevMagazine site you want to increase the number of visitors that fill in a form to subscribe to your newsletter and thus increase engagement and collect contact details.
3. Choose on the page you want to optimize.
   For example, the Subscribe call-to-action button is situated on your Home page, so you are testing the original Home page against its variations.
4. A/B test your hypothesis.
   You design variations of your original Home page and run the test for an amount of time, relevant to your goal and page traffic.
5. Measure results and analyze the data.
   In Digital Experience Cloud, examine test results and, based on a combination of goal completion rate and statistical significance of results, pick a winner and variation that delivered the best performance, according to your needs.