Personalizing an experience is not easy. Personalization cannot be treated just as a feature and can require significant effort. The good news is that personalization doesn’t have to be a manual process. When using a manual process for personalization (which typically involves writing a bunch of rules) it’s impossible to scale.
Whether you have to deal with large catalogs with ever growing number of items and SKUs, multidimensional spaces describing product (or other content) features, defining and managing a large number of explicit rules quickly becomes ineffective and overwhelming. In addition, manual processes are slow in responding to changes in the market place and customer trends. Retailers are learning that in order to keep up with growing catalogs and SKUs (not to mention the competition) an automated process is critical to success.
Based on machine learning, automated personalization involves techniques and algorithms that identify behavioral patterns. These learned behaviors are used to infer user’s intent and to guide both the user’s discovery process as well as to offer a better response to the user actions in a given moment.
Ever wondered why traffic is heavier in some aisles than others, or why retailers place certain products at eye level instead of below eye-level? The truth is that our behaviors, needs and interests are actually more in the same than they are different. It is the patterns, or combinations, of those interests and behaviors that make us relatively unique. And, I say “relatively” because even within those patterns, there are lots of people just like us. eCommerce personalization engines recognize this to some level, but we are still learning. You can learn more about the human need for being “known” in combination with automated personalization in the free paper, download it now.