Quite a few years ago I came across a quote (the source of which I can’t now track down) that read, “The information revolution is over. Information won.” Back then we were only dealing with the emergence of e-commerce and the recognition that there was value in the “information exhaust” from business activities. In the ensuing 20 years, we have seen an information torrent turn into a flood and then a deluge — a deluge that threatens to swamp our ability to collect, analyze, store and maintain (as well as make intelligent use of) all the data that’s available to us.
At about the same time I came across that quote, I was working with a very large global retailer that was a pioneer in using information to improve its operations. Retail is a tough business, often with razor thin margins. But supply-chain logistics is one area where data can really make a difference.
How? It used to be a common saying in logistics that “uncertainty is the mother of inventory” — excess inventory ties up capital and erodes margins when it has to be cleared out via “markdowns.” On the other hand, running out of products that are in demand lowers revenues (you can’t sell what you don’t have) and reduces customer satisfaction. In retail, customers generally have choices and you may never get back the customer whose demand you fail to meet. Getting this balance right used to be an art form. The retailer I was working with wanted to make it a science.
By collecting point-of-sale (and point-of-service) data in close to real time (initially several times a day, eventually continuously), the retailer could see which items were moving faster or slower than expected; which inventory was available in the stores, on trucks in transit and in distribution centers; and where that inventory should be sent to best match predicted demand. “Yield optimization” was the name of the game back then (often called “efficient customer response”). Best-in-class capability was worth a couple of additional points of margin as well as improved customer loyalty.
It took several years and a lot of investment (and some very smart people) to get the system to be that good. By the time it was done, if the retailer knew who you were when you entered one of their stores it could predict with better than 80 percent accuracy what you were going to buy. With that insight, the retailer could target promotions and use cross- and up-selling programs to improve its yield even further.
There were however, three issues with this.
Note that there are a couple of ironies here. The retailer was so good at operating its physical retail business that there was little incentive to jump online (it preferred to add more categories to its stores, leveraging its scale and its core merchandizing and supply-chain capabilities). Online, the retailer would have known who its customers where when they showed up — something that’s very hard to do in the physical world. And online customers (at least back then) knew they weren’t going to get instant gratification — they knew they would have to wait for a purchase to be delivered. Thus the retailer could have gained considerable flexibility in order fulfillment by going online.
So what are the lessons here?
Data helps, but only if it’s applied to the right problem seen in the right light. Otherwise you’re just paying for (a lot of) something you can’t use.
John Parkinson is an affiliate partner at Waterstone Management Group in Chicago. He has been a global business and technology executive and a strategist for more than 35 years.