A/B testing is a simple way to test changes to your page against the current design and determine which ones produce positive results. In its most simple form A/B testing is done by sending out the same webpage with different designs to two different user groups, it is then determined which of the user groups has a higher conversion rate. A/B testing is therefore an essential part of conversion rate optimisation and is important for anyone who has an Ecommerce store or needs to gather leads through their website.
1. Technology is your friend: Bucket testing is very easy now thanks to the Internet. An example of an email test is to create the same offer to two groups of people using two codes. The code used more times indicates the more appealing offer. Use the multitude of software and information available before you start your conversion optimisation process. Even though technology is important a good way of testing web page design is by asking real customers to do it in front of you physically and then give you feedback afterwards.
2. Eliminate friction: Friction is the thing slowing down the process of your customer landing on your page/website/ad and the actual conversion. Locating friction is the most essential element of successful A/B testing. An example of friction are long forms that want browsers to enter large amounts of unnecessary information. Shorten these forms to the bare minimum in order to achieve optimum conversions
3. Where is your Call to Action: a call to action is essential to gathering leads on line. Call to actions can be a ‘buy now button’ or a ‘sign up for our newsletter’. Make sure you test your call to action design, position and wording often
4. Know your customer base: Every company has target customers based on its services. Your A/B testing will be much easier if you know where you are starting from.
5. Don’t make lackluster ads or pages: There is no such thing as the better of two evils – they are all evil. Don’t undercut one test for any reason.
6. Beware of multivariate testing: A/B testing suggests that people are either choosing between options “A” or “B.” You can use multivariate testing but you will require additional testing in order to make a conclusion.
7. Use A/B testing on things other than ads: If possible test everything you can instead of focusing on pages or ads.
8. Change feature placement: You can change the placement of graphics text and layouts for more accurate A/B testing. A simple change might cause a big change.
9. Study SEO options for A/B testing: You should use SEO to reach your target group in order to get more accurate results.
10. Read both negative and positive results: You should know that not all negative results are detrimental to your site.
12. Understand your results: A/B testing is complex but you should take time to learn how to interpret results. Never stop testing – customer preferences are always changing not to mention you can never have a perfectly designed webpage or advert. There is always room for improvement