How Sitecore’s machine learning amplifies your test results

When you run tests for page optimization, you end up with clear results that tell you which page layout, content, or images worked best with all your site visitors. But what’s missing in regular tests is an important nuance that could help you drive better results for your company.

Suppose your company sells insurance online. When visitors come to your homepage, they click information for a specific car insurance product, and you show them variants A and B of the same page. After running this test for a while, you see that page A is leading more people to request an insurance quote. So you should just scrap page B since it’s not working as well, right?

Traditionally, the answer has been yes. But Sitecore’s machine learning feature digs deeper into your test results to show you that things aren’t so black and white. With machine learning, Sitecore analyzes your test results and clusters your visitors into audience segments — revealing that page B actually works a lot better with millennials. This valuable insight means you’ll keep that page and use it for specific targeting, but you’ll show page A to baby boomers, who make up the majority of your customer base.

The result? Better optimization and performance for all your potential customers.

One of the major benefits of using machine learning with testing and optimization is the ability to effortlessly detect new audience segments. Let’s take another example: imagine you’re an insurance company optimizing your car insurance product page.

While you were busy doing other things, Sitecore was analyzing your test results and clustering your visitors in a meaningful way. Let’s say you’re located in New York City and have always focused your efforts on marketing to customers in the surrounding areas. But thanks to Sitecore’s machine learning analysis, your insurance company now knows some of your website visitors are from California, and others are also interested in RVs. Suddenly, you’ve discovered untapped opportunities: new regions and potential cross-sales.

All it takes is a bit more personalization to target these visitors — new potential customers you didn’t even know you had.

When to start with personalization

If you already know you have audiences with very specific intent and needs, you should start personalizing directly for them rather than testing your content. That’s because you know what they are looking for and can provide a direct path for them to get what they want.

Also, if you have different audience segments in the same stage of a lifecycle, you should start with personalization as well. By creating content that’s personalized to each segment in that lifecycle stage, you’ll provide great value for these visitors and achieve higher engagement. You can then run tests to validate that your experiences are getting the results you expect.

When to start with testing and optimization

If you don’t know who your specific audiences are, always start with testing. Your testing will let you optimize components for overall site experience, page content, and more. And with Sitecore’s machine learning capabilities, any test you run will help you identify your audiences, which you can then use to personalize your content.