Not long after Christmas, shoppers will descend on malls to scoop up marked down sweaters, jeans, toys and ski equipment. You would think that with all of the data available to big retailers, they would know how to predict what will sell and what won’t. Clearly, though, many still don’t know how to do so accurately—and it hurts their bottom line.
That is changing, however, thanks to predictive analytics. The technology allows retailers to use statistical models and data they have gathered to foresee what consumers are likely to buy. A retailer who sells high-end coffee makers where coffee is prepared in K-cups could, for instance, determine exactly which customers bought the coffee makers in the past and when they are likely to come back to the store to buy more K-cups—or figure out which ones would come back sooner if the store ran a sale on them.
Some retailers are using predictive analytics on a large scale. URBN—which includes brands such as Urban Outfitters, Anthropologie, and Free People—uses predictive analytics technology extensively to make better decisions about how to stock its brick-and-mortar stores and what to sell online.
Predictive analytics help retailers make better business decisions while also providing more tailored customer experiences.
However, predictive analytics aren’t always 100% accurate, as we saw with Google Flu Trends. Initially, it seemed like Google’s algorithm could predict where the flu would break out, based on search terms being entered, but later those forecasts turned out to be inaccurate, possibly because there were other reasons, like news reports, that made people search for information on the flu. There are similar gaps in the information that predictive analytics can provide to retailers.
Nonetheless, predictive analytics is only going to get better as technology improves and more data becomes accessible. The next frontier for retailers will be finding the right incentives to persuade consumers to make privacy tradeoffs. Retailers using predictive analytics currently rely on data on customers’ past purchases, online searches, past shopping patterns, and the amount they have been willing to pay for products in the past to predict how likely they are to buy at certain price points.
To elicit more information from customers, merchants will have to get creative and offer something that consumers truly value in exchange for access to their data. Simply offering discounts, even big ones, may not be enough of a tradeoff for privacy.
The next frontier for retailers will be finding the right incentives to persuade consumers to make privacy tradeoffs. Discounts, even big ones, may not be enough.
Some companies are embracing the creative challenge. Burberry, which now sells its fashions straight off the runway to online customers, mitigates the risk of manufacturing too many versions of a particular design—or too few. They use predictive analytics data gathered from fans of its social media pages, such as Instagram and Snapchat to indicate appetite. Disney has patented a system that will track guests by their shoes so it can create a customized experience for them at its amusement parks.
If such experiments with predictive analytics allow companies to create an optimized and effortless experience that consumers love—and merchants don’t abuse their access to information they gather—many consumers are likely to accept the tradeoffs in privacy they must make to enjoy experiences like shopping at the mall or their vacation even more than they already do.
Read about our vision of predictive retail in “Harvard Business Review”.
We’ll be talking about this and other cutting-edge topics with some of the best minds in retail at our invite-only Narvar Summit on February 9. Follow the conversation on Twitter at #narvarsummit.