The analytics world has been using the term “journey” as the new buzz word to describe the forward looking, predictive aspect of what a customer could buy. However, are people really looking at the journey or are they simply looking at the next immediate step in that customer’s path?
The Long Term Trajectory
Most predictive analytic platforms do predictions by some form of affinity modeling to try to predict what may be the next item a person may purchase. These platforms do so by a variety of methods, such as by category, by most recent purchase or even by demographic profile, but in each case they focus only on the next step.
Using statistical models, it is possible to come up with “next best action” type suggestions that could be appealing to a consumer in the near-term. However, what is missing in this process is factoring in not just the item that has immediate appeal but a way to model what is the long term path and trajectory of an individual–not just the next step, but the one after that, and the one after. I should be able to judge the appetite of a person to try new things and offer them products that meet that “hunger.” I should then be able to guide and nudge a person in the direction I want them to take. I should be able to send them on a journey based on how fast they are willing to go and the general direction they are headed in order to maximize their value to my business and also meet and exceed their expectations.
How do we Influence this Journey?
In order to create this guided multi-step map that takes consumers through a journey, analytics must do more than just offer near-term recommendations. Analytics have to create a multi-dimensional view of each consumer, layering in products to that view and then trigger appropriate suggestions and recommendations along the way to guide a consumer in the direction I want them to go. In order to be focused on the journey, you have to predict journeys and pathways for each consumer. This is not simply doing affinity type modeling that analysts have been doing for years. The paths themselves have to adjust based on the business goal-set. For example, is the business looking to be revenue focused? Are they looking to increase share of wallet? Are they looking to achieve more repeat purchases, or just engagement overall with the portfolio? Each one of these business goals impacts the journey that you want your consumer to embark upon, and appropriately mapping this journey becomes fundamental.
This becomes more of a question of painting the picture of each consumer’s pathway. In mathematical terms, we look at the “geometry” of the space. At Coherent Path we use a type of geometry called hyperbolic geometry to predict these journeys. Large and small retailers have been using our analytics platform to improve their business – they tell us what they are looking to achieve and then, using hyperbolic geometry, our platform predicts the journeys for millions of their consumers. It is one thing to say that you are focused on the journey that your customer is pursuing, but it is quite another to have a methodology to effectively map it and a platform that automates the process.
So, the next time people bring up journey in an analytics conversation, I would ask them if they are truly looking at predictions that go multiple steps out rather than simply labelling today’s affinity modeling and next-best-action type recommendations as “journeys”.