Predictive e-shopper analysis software boosts marketing ROI
Gilbert Systems has launched a new product that forecasts online consumer purchasing behaviour. The 'Proclivity' software aims to allow retailers to identify customers with a high purchasing likelihood for a specific product, and to use that information to tailor their marketing campaigns for greater profitability.
The Proclivity system, which helps predict not only what customers are likely to buy but also when and at what price, is currently running in pilot programmes with a number of online retailers and publishers.
Online activity analysis The company describes the system as an "adaptive platform" that analyses each customer's online activity and identifies their interests in various products to help predict future purchasing behaviour. Proclivity uses probability indices to identify customers that are likely to buy a particular product, at a specific price point, and within a defined period of time.
The predictive engine can adapt to variables such as product categories, price points, and even seasonal product cycles, and it can be applied to a variety of e-commerce sectors, including retail, travel, finance and media.
Marketing decision support According to Sheldon Gilbert, CEO for Gilbert Systems, the system supports and informs users' marketing decisions, identifying customers with high purchasing likelihoods and forecasting the return on investment (ROI) for campaigns.
One of the major points of differentiation for retailers using the system, though, is that the platform can produce unusually timely and relevant promotion e-mail messages for customers, on a one-to-one basis, because it has already predicted which products or services are most likely to be relevant to them at the time. As a result, Gilbert says, order conversion rates and sales productivity will increase while customer attrition will decrease.
Impressive results Gilbert reports that some of its pilot programme clients have seen thirteen-fold increases in online order conversion rates, up to 300% increases in e-mail open and click-through rates, and up to 600% increases in campaign productivity (based on revenue per e-mail).
"Most companies try to use generic demographic, purchase, or survey data to anticipate consumer shopping preferences. But we use data that is self-generated by each individual consumer to make more relevant and accurate predictions. This provides finer resolution into the true tastes, preferences and shopping habits of consumers for individual products," said Gilbert.
How it works As a customer interacts with the company web site, clicking on objects, searching for information, enlarging a picture, placing an object in a shopping cart, making a purchase, or any of a number of other noteworthy and pre-defined actions, the customer leaves a distinct digital "trail of interest". This provides clues as to the types of products or services that the customer is interested in.
Each step along that trail is examined by Proclivity and used to predict the customer's future shopping activity and purchasing habits. Every step the customer takes has a value that correlates to the likelihood of future actions, and Proclivity calculates the value of these steps by analysing correlations at the individual customer level and using aggregated behaviours from all of the site's customers.
At the same time, Proclivity examines the behaviour of content or products on the website. As prices fluctuate, incentives are offered and product attributes shift, Proclivity probes the effect of these changes on customer interactions with the product.