It is certainly actionable to find one customer segment, and personalize to that. Just find the common traits for that group – maybe their location, or if they have purchased for a certain amount – analyze their behavior, and make the appropriate variation just for them. The tools are all there, they're easy to use, and performed once it's not a major project at all. The thing with effective personalization however is that it needs to done near continuously, with constantly evolving segments. Personalization is a fluid, not a static process.

Most marketers, including myself and my colleagues, wear many hats: brand ambassadors, lead generators, content marketers, digital analysts, customer evangelists, and the list goes on. There are many responsibilities, but only 24 hours in a day. Because personalization today depends on, as time goes by, an intricate set of rules that have been manually set up (at one point), it's often the first thing on a marketer's list to go. In our experience, automation is the only way to make personalization scale, and work effectively in the real world. The trouble is that so far, automation has required an even bigger investment not only in technology, but elaborate configuration to get the machine going. We're about to change that.

We call it "autonomous personalization," and it consists of three connected parts:

• A data store, that gathers intelligence from visitors' actions – built for massive scale.
• Machine learning algorithms – sometimes labelled "predictive analytics" – that find correlations and groups within the data from thousands (or millions) of visitors.
• And finally, a tool that lets marketers and merchandizers simply select where to display personalized content, and let them see the improvements in conversions, reduction in bounce rate, or whatever their goals - in real-time.

This set of connected technologies will not only enhance current personalization and one-to-one targeting efforts that are now quite common (such as product recommendations), but make personalization accessible to a number of industries and use cases that previously haven't deployed it at scale. For example, B2B companies can serve personalized content based on their visitors' individual and collective behavior, providing buyers with content that they are most likely to need.

The complexity of handling arbitrary content, rather than a well-structured set of products, has so far proved too difficult for technology to manage. We look forward to combining state-of-the-art algorithms with a deep understanding of content to help organizations and brands personalize on a continuous basis - autonomously, and ultimately provide their customers with a truly individualized experience.