Every time we use a credit or debit card or a service like Apple Pay or PayPal to make a purchase we are sharing data about ourselves with the retailer concerned. Individual purchases may not reveal much but over time the data on multiple transactions can be used in sometimes quite alarming ways.
An example of this was US retail chain Target’s now infamous pregnancy prediction project. The chain used advanced data analytics techniques to track female customers’ purchasing patterns and assign a “pregnancy prediction score”. Customers who bought certain lotions and products associated with pregnancy naturally received higher scores.
The retailer used the resulting information to send vouchers for baby clothes and other related infant care products to the customers in question – often with almost spooky results. In one case they received a complaint from a male caller who was outraged that they had targeted his teenage daughter with these vouchers. He later apologised and confirmed that his daughter was indeed pregnant but hadn’t told her family yet.
Extremes
This may be taking it to extremes but retailers and payment service providers are constantly looking for new ways to use the data gathered by payment systems both to improve the bottom line and make life better for customers.
One example of this is the use of point-of-sale systems in restaurants to improve efficiency and enhance the customer experience. "The apps which can be loaded on to our Talech systems allow restaurants to analyse what customers are buying and when they are buying it," says Eric Hogan, country manager with Elavon.
“That’s where we see business going. There is so much data going into one transaction that can be used to make decisions about the business.”
In the restaurant example, the apps loaded on to the device will integrate with stock control and other systems running in the background. At the end of the day they will generate reports on best- and worst-selling items which can be used for the design of menus in future.
The reports will also show if there are particular demand patterns which indicate a need for extra staff in the kitchen or on the floor waiting on tables. Poor-selling items which end up in the bin can be discontinued and more popular ranges can be extended.
Next wave
The next wave could see part-filled shopping trollies awaiting us in supermarkets. Shoppers alert the retailer to their impending visit and a trolley is filled with likely purchases based on previous shopping patterns, the time of year, the day of the week and so on. The shopper simply takes out the items they don’t want and puts them in a basket near the entrance and proceeds to add those that they do need.
And that won't be the end of it by any means. Writing in Forbes magazine two years ago, big data specialist Bernard Marr envisaged a time when drones will arrive at people's houses delivering goods they haven't even ordered yet.
Maybe this isn't too far-fetched, it is after all the same model used by the old mail order book and record clubs which existed before we had smartphones and Amazon.