How do I setup and run an A/B test?

What is A/B testing?

A/B testing (also known as split testing) is a process of showing two or more variants of the same Campaign to different segments of website visitors, so you can compare which variant drives more conversions.

Why is it useful?

A/B testing is a great way to fine-tune your design and message for the best possible results. 

How to set up A/B testing?

  1. First, log in to your OptiMonk account at https://app.optimonk.com/login/en

  2. Select Campaigns on the left, then select the Campaign you wish to use for the A/B test:

    mceclip0.png

  3. You will need to create at least two versions of your campaign. To add a new variant, click on Add new variant:

    mceclip1.png

  4. You will be able to A/B test right after Activating the new variant, by clicking on the red toggle:

    mceclip2.png

What should I change to A/B test my popup?

OptiMonk gives you the freedom to create different variants according to your choice. Your variants can also be completely different. 

You can edit the design of the different variants by clicking on their name:

mceclip3.png

We recommend to change the elements of the popup in the following order:

  • Headline
  • Subheading
  • Button text
  • Data asked
  • Colors
  • Other texts and descriptions

 In addition to the content of the popup, you should continuously refine your main ‘offer’.

Maybe your customers would react ten times better to a totally different offer/communication. A discount might get you more conversions than a free ebook. Every site has its own visitors with their own set of needs.

There’s only one way to know which popup will work best for your visitors: by testing it!

If you’d like to know more, read our case study on how an e-commerce store used A/B testing to have better results:

https://www.optimonk.com/blog/case-study-how-an-ecommerce-store-reduced-their-cart-abandonment-rate-using-onsite-retargeting/#.Wyub8EiFOM8

Please note: We do not recommend to make too many changes because then you won’t be able to determine why each variant is performing better or worse than the others.

A sensible way to get it started is to make a change to one element, test it, and then make a change to the other elements as well, choosing the best performing variant each time.

Have more questions? Submit a request

Comments