How to Improve Website Personalization with A/B Testing


If you’ve visited your dashboard recently you might have noticed the newest functionality in our Audience Targeting rules creator – the ability to create A/B tests with multiple rules to see which perform best (learn how to set up your first test!). Paired with the ability to create sophisticated and customized messages, Audience Targeting is now better than ever. But why are A/B testing and personalization important? Let’s find out.

If you run a website, you probably already understand the independent value of both A/B testing and personalization. What you might not have considered, however, is how those two factors can be even more effective if you use them together. First, let’s take a look at how each is vital to a successful website in its own right.

A/B Testing Helps Refine Strategy

A/B testing—also known as split testing—is delivering two different versions of the same web page to visitors to determine which one performs better. The results can then be used to inform website redesigns, refine new tools and functionality, and most importantly, improve conversions. Here are just a few things that you can A/B test:

  • Calls-to-action: Strong calls to action are one of the most important tools for driving both clicks and conversions, so figuring out which ones will perform the best is vital. You can test colors, copy and even text vs. button to see which drives the most sign-ups and sales with your audience.
  • Images: People respond differently to visual elements, and the ones you choose make a big difference for conversions. Try testing product shots vs. people, static images vs. videos, or even image sizes.
  • Copy: Viral success sites like Upworthy and Buzzfeed do vigorous A/B headline testing to discover what works. If you spend significant time on social media networks, you’ll know that it pays off for them. In addition to headlines, you can test tone, SEO copy, and share copy to see which pays off.

The list of site elements that can benefit from A/B testing is long and constantly evolving. You can use it to test virtually anything, but it’s important to not go overboard. If you try to test too much at once, you might not have a large enough dataset to pull anything meaningful. Focus on one test at a time, and when you’ve exhausted it for learnings, move on to the next. Over time, you’ll have a winning formula for delivering web pages that work.

Personalization Drives Conversions

As more websites begin offering personalized experiences, more visitors will come to expect it. If you visit a website populated with products, services, and content that you’re interested in, you’ll be more likely to continue on the path to purchase or sign up.

Here are just a few ways to personalize content to your visitors:

  • Target different messages to new vs. returning visitors. If you’re a new visitor, you might be interested in learning more about the company and the products or services they offer. A returning visitor may be ready to begin the path to purchase. A loyal customer may be stopping by to see what’s new. Targeting specific messages to each of these groups will help them to engage with your brand in a way that makes sense to them more quickly.
  • Offer different experiences for mobile and desktop users. People are increasingly using smartphones to browse the internet, and they need a different website experience than those who are visiting your website on desktop computers. Make sure your pages are easy to navigate and effective on any device your visitors may be using.
  • Create custom landing pages. Make campaigns more effective by creating landing pages based on user demographics and behavior. You can target based on where they came from, where they’re located and whether they’ve visited your website, just to name a few possible parameters. AddThis Audience Targeting makes it easy to personalize pages and content based on the factors that are important to your business.

A/B Testing and Personalization Make a Great Team

A/B testing and personalization are both effective on their own, but if used together, the possibilities are endless! Consider this: you’re A/B testing shows you that 65% of your audience wants X, while the other 35% wants Y. If you then decide to go with X across the board, you’re potentially throwing away more than one third of your audience!

Use analytics to figure out what makes a user more likely to be in group X or Y. Is it the website they came from? Where they’re located? Whether they’re new or returning visitors? Once you know who each audience is, segment them and do more A/B testing.

Rinse, lather, repeat—and watch your conversions grow.

The important thing to remember about A/B testing and personalization is that your work is never done. Seasons change, trends change, and audiences evolve. Continually trying new approaches, testing their effectiveness, and refining your strategy based on what you learn will ensure that you’re meeting the needs of your website visitors. If you do that, they’ll reward you with traffic, sign-ups, sales and most importantly, customer loyalty.

To get started running your own A/B test, simply login to the dashboard, create an Audience Targeting rule and you will be given instructions on how to launch the test!

Visit the AddThis Academy and check out the newest article “How to Run A/B Tests With Audience Targeting” by our very own Software Engineer, Josh Horwitz.