Press

« View all press releases

Retail Success is About Who’s Working When.

Source

by Serguei Netessine - Harvard Business Review – December 6, 2011

Along with the growth in scale of leading retailers in the 20th century came a growing attitude toward the people working in the stores: they were a cost to be minimized. Sam Walton, the founder of the biggest retailer in the world, summed it up in his book Made in America: “No matter how you slice it in the retail business, payroll is one of the most important parts of overhead, and overhead is one of the most crucial things you have to fight to maintain your profit margins. That was true then, and it is still true today.” But recent research by my colleagues and I suggests that retailers are thinking far too simplistically about the cost and potential value of their workforces.

Let’s start with stockouts, a problem most big retailers are highly attuned to; they know that when a customer arrives intending to make a purchase and finds the shelf picked clean of the desired item, the store not only loses a sale but also damages the likelihood of that customer’s returning. Fixated on that challenge, retail chains have invested heavily in sophisticated inventory management systems. Yet at one large retail chain we studied (pdf), those systems didn’t seem to be doing the trick. When we analyzed results of a customer survey, we found that nearly 20% of the products they wanted to buy were out of stock. This was despite the fact that, according to the inventory management system, only 2-3% of items ever ran out before being replenished. It wasn’t that the system’s numbers were wrong. The problem was that customers couldn’t find what they were looking for—and without a store associate to help, they left empty-handed.

Saving sales by pointing to merchandise locations is just one of the ways that store employees facilitate the sales process and perform a very important role. But in large retail enterprises, it’s easy for managers to ignore the details of sales floor interactions and opt for large-scale, broad-brush solutions to the challenge of staffing. Most simply set targets for store staffing levels they must maintain over time (mandating, for example, that the cost of labor cannot exceed 10% of sales), and then apply that level across the board. At best, they vary staffing levels based on sales forecasts. Almost invariably, such overall targets lead to a situation where some stores are overstaffed while others are understaffed.

Given today’s technology available for data acquisition as well as new developments in analytics, it is possible to do much better than this. Rather than simply predicting what volume of merchandise will sell in a certain period and scheduling more or fewer labor hours accordingly, it is possible to observe the actual flow of customers through stores and make adjustments—even in real time by moving additional employees to the sales floor or redeploying them to higher-traffic areas. Our studies have shown the wisdom of this: stores that manage labor levels in light of store traffic rather than sales forecasts achieve substantial sales increases without extra costs.

Even better results come when retailers recognize that their workforces are not just homogeneous pools of labor to draw on. The most innovative employee managers we know use business analytics to understand the differences in how individual store associates perform. When these retailers make dynamic adjustments, therefore, they are not only deciding how many but who in particular to move to a sales floor. Ann Taylor, the women’s clothing retailer, has been a pioneer in tying staff scheduling to the past performance of sales associates: its best salespeople get first choice of times to work and more schedule flexibility. It’s a capability that is also finding its way to other business sectors. Call center companies are increasingly capitalizing on their ability to track individual sales to match the most effective people with the right opportunities. A Boston-based startup, Objective Logistics, has just received a round of funding from Google Ventures, Atlas Ventures, and a few other investors to bring performance-based scheduling software to restaurants. (Full disclosure: I am an advisor to the company.) The system would offer up the best times to work, like Friday and Saturday dinner times when bigger orders generate higher tips, to the most productive waiters.

Scheduling, of course, is not the only aspect of labor management that retailers must get right. Once they’re at the counters and in the aisles, employees also need the knowledge of how to interact effectively with customers. This is a hard thing to understand—let alone teach—and an area where managers find simplistic solutions just as tempting. Take, for example, the edict some retailers have issued that each customer must be greeted with a smile at the entrance of the store. (Our research has found essentially no relationship between greeting the customers and store sales.) To better understand the nature and the outcomes of customer-employee interactions, a handful of companies are currently experimenting with analysis of in-store video streams. (Note that privacy concerns hamper such video analysis in the US. My experience of it has been in South America, Asia, and Europe, in collaboration with Bravo Lucy, a Norwegian and Indian business analytics firm.)

Scheduling, however, is an area where the tools exist to manage better today—and where the evidence is clear that managing simplistically can send a retailer into a dangerous downward spiral. In a recent study, colleagues and I used data from major retail chains and found a common trend of low forecasted sales for a weekend resulting in stores staffed too thinly to provide adequate service for the customers who actually showed up. The lost potential was evident in the long lines observed at checkouts and the poorly stocked shelves; undoubtedly many customers left the store without buying. But for any manager looking only at receipts and staffing, the end of the day brought vindication: sales, indeed, were low. Isn’t it remarkable how prophecies can fulfill themselves?