A/B testing is a method of testing that involves forming two or more variants of a website and then determining which one performs better. It is most commonly used to refine web design and, more importantly, boost conversion rates.

The importance of personalization

As more and more websites offer personalized experiences, visitors are coming to demand and expect them. When customers visit a website with content, products or services that are relevant and interesting to them personally, they are more likely to be engaged by making a purchase or signing up for content.


A/B testing for website personalization 

A/B testing and website personalization go hand in hand. Using A/B testing, you can determine what works to charm customers and what does not. Further, by identifying the customers that are not engaged, you can dig deeper to try to understand why and customize the experience of those customers as well.

Here are some ways you can use A/B testing to improve web personalization:

1. New customers and returning customers – A/B test different offerings to new customers against returning customers. New customers want to see what your concept is and what your website is all about. They are in the getting-to-know-you phase. Returning customers may just be checking out updates on past interactions or what’s new. Delivering customized messages to each kind of visitor ensures you engage both audiences in a way that is meaningful to them.

2. Mobile and desktop users – A/B test different experiences for desktop and mobile users. Users are increasingly connected on their smartphones and the functionality of these devices is different. Ensuring your website translates well to mobile viewing is essential in engaging with these audiences as well.

3. Time spent on page – When users feel that a website does not address them personally, they are more likely to leave without doing much. A/B test variants to evaluate what results in more time spent on the website.

4. Call-to-action buttons (CTA) – Personalize various CTA buttons based on the desired end goal for your customers. A/B test whether placing the button in the middle of the page results in higher conversions than placing it at the bottom. Other aspects you can test are colour and wording. With different customers, you should offer CTA buttons that make sense to them. For example, new users should be presented with an easy to find signup button. If they only see the sign-in button that applies to existing users, they are unlikely to go hunting for the path to register. At the same time, the sign in button should be easily accessible to returning users.

5. Video versus image versus text – A/B test the different pages of your website to determine whether users respond better to text, video or images. This will differ on different web pages. For example, a video that demonstrates the features of a product may be more effective when your customers are not businesses. However, businesses may prefer a list of specifications for their records or for their internal discussions. By A/B testing a video, you could judge customer curiosity. A/B test a hyperlinked list of specifications beside the video and monitor click rates and who they are made by. Are businesses as engaged by the video as other customers? This information could be useful in devising the perfect landing page to amplify customer engagement.

6. Registration methodsSignup forms used to long and arduous to fill. Nowadays, the Internet is populated with a wide variety of people with varying computer experience. Shopping, social media and payments are increasingly undertaken online by individuals and businesses alike. Users are keen on an experience that is simple and intuitive. When signing up or creating an account, users have come to experience a form that is simple, minimal and requires only the information that is absolutely necessary. They will usually prefer one website over another based on usability alone. This is true particularly of individuals but businesses are also considering user experience in view of gains in productivity and efficiency. A/B test single step versus multi step signups, single page layouts versus multi page layouts etc.

7. Schemes and offers – Has your website considered the option of free trials? What about returns and money-back guarantees? Knowing what your audience expects is key. For example, in the software industry, customers have come to expect a free trial and even a demonstration. Some even offer training programs. A/B test customer response to 14-day free trials versus 30-day free trials. For example, if you are trying to engage large corporations, A/B test various competitive pricing schemes and a convincing landing page. A/B test key words and phrases in your schemes to see what has an impact.

8. User location – Delivering content based on the users’ location by detecting country or zip code can help you deliver personalized content to your customers. Deliver local content, provide shipping details and invite visitors to events in and around their city. A/B test the modes of engagement that are most successful. For example, test mode of delivery of shipping details. When a local visitor enters your website, A/B test whether a short notification on the home screen about shipping to their country is more effective than providing shipping details at the time of making the purchase.

9. Identify visitors’ channels of access – Are visitors accessing your website by Twitter, by Facebook, through a search engine or by way of the URL? If your visitors arrive via Twitter, for example, a little digging around their profile will tell you what they are interested in. That way, you can A/B test various content to see what is engaging. If visitors arrive via a search engine, you can evaluate their search results and the time spent on your page. This will tell you whether they find what they are looking for. You can also present them with related articles to maximize the value they derive from your page for their search results. With a little A/B testing and experimentation, your page can be optimized for various classes of users.

Web personalization is a continuous process and A/B testing can be an excellent tool to determine the effectiveness of various personalization strategies. Both web personalization and A/B testing are continuous processes. When you have completed A/B testing of every single element, restart. See if your tactics are still effective. There will always be room to improve.

Image Credit: https://vwo.com/ab-testing/