Blog
    Opinion
  • Predict conversion rate for each visitor with machine learning

Unless helps websites predict conversion rates per visitor with machine learning for more sales and less product returns.

*This article and similar versions of it were recently featured in various local news publications (in Dutch) such as Twinkle, FONK, Emerce, Marketing Tribune and Marketing Report. We translated the content to English so that it is accesible to our international audience as well.

Unless, a personalization platform for websites, has launched a new application that can be used to predict conversion rates per visitor. The app is available for use on any website. The company’s technology shows highly personalized content to visitors, resulting in greater ROI.

People have a short attention span. This has big consequences for businesses on the web. On average, only 2% of website visitors convert. "If you could appeal to every visitor in a personal way, your conversion rates would improve dramatically,” says Sander Nagtegaal, CEO at Unless. “But you would have to know what kind of person each visitor is. This has just become a lot easier with machine learning.”

More sales and less product returns

Unless’ service can be integrated with any website. Users set their own business goals to measure success. “If you run an e-commerce shop and you sell shoes, every pair sold counts as a goal completion,” Nagtegaal explains. “You keep track of risks, such as product returns, in the same way. By recording these events, you can start building visitor segments based on pattern recognition - so you can better understand who is most likely to convert.”

Unless’ new algorithms predict the likelihood of conversion for every visitor. This data can then be used to personalize your website. The user experience can be customized for each target group using the personalization platform’s editor. Think of customized headings, text, images or buttons that appeal to each target group in the best possible way. Unless ensures that every website visitor sees a unique version of the site, crafted especially for them.

Purchases from a target group that has a high conversion value can be stimulated, while those with a lower conversion value can be encouraged less. This can be valuable to prevent product returns. “Driving more sales is nice, but reducing costs by avoiding return shipments is also important,” says Nagtegaal. “This can make the difference between success and premature bankruptcy, especially for new e-commerce businesses.”

Machine learning and artificial intelligence

Unless is betting big on machine learning and artificial intelligence. “Very soon, we’ll be able to predict not only who is more likely to convert on a website, but also what works best to increase their attention span,” adds Nagtegaal. “Our goal is to make websites more human, and we’re getting closer and closer to reaching that goal.”

In two years’ time, Unless will be able to automatically adjust the user experience based on the website visitor’s background, behavior and even mood. “A longer attention span and better content will benefit businesses and consumers alike,” Nagtegaal concludes. “Consumers will be less frustrated with content that is not relevant to them, and businesses will see higher conversion rates online.”