Post Your Comment
Rocket Science Retailing: A Practical Guide
When it comes to implementing science in retailing, the 'missionary' is more important than the scientist
Whether you sell widgets, designer fashions, or life-saving drugs, mastering the art and science of better analytics can set you ahead of your competitors, according to HBS professor Ananth Raman and Wharton professor Marshall Fisher.
Raman, who specializes in operations management, is the UPS Foundation Professor of Business Logistics at Harvard Business School, where he teaches courses in supply chain management, service operations, and the investor's perspective on operations to MBA students and executive education participants. Fisher, the UPS Professor of Operations and Information Management at the Wharton School at the University of Pennsylvania, is codirector of the Fishman-Davidson Center for Service and Operations Management.
An excerpt from their book follows our interview.
Martha Lagace: In a nutshell, what is rocket science retailing?
Ananth Raman: It means that retailers should
* Use the data generated at your stores to understand customers and their needs deeply.
* Develop the ability to respond to this understanding with better-tailored assortments, replenishment of the hits, and timely markdowns on what is left over.
* Execute well, especially at the stores. Attend to data inaccuracy and placement of products within stores.
* Align incentives within your organization and in the supply chain.
* Use technology judiciously and pay attention to emerging new technologies, whose value might still not be apparent.
* Explain the changes you are making to your investors.
Q: In The New Science of Retailing, you and Marshall Fisher write that "well-designed incentives are a necessary condition for rocket science retailing to work." What are some typical sources of incentive misalignment? How can incentive misalignment be reduced or overcome?
A: We frequently see perverse incentive misalignment within organizations and in the supply chain. In the book, we identify three reasons for perverse incentive misalignment:
1. Incentives exist to induce specific behavior. Managers who design incentives often are not entirely clear on the behavior they would like to induce.
2. Managers who design incentives often do not have a sufficiently deep understanding of operational details.
3. Most operations—both within the firm and in the supply chain—are rife with "hidden" information and action.
To align incentives better, managers should first acknowledge the role of incentives in guiding behavior. Then, they should develop a deep understanding of their operations and the behavior they want to induce. Finally, they should trace the incentive misalignment to hidden action or information and use contract-based, information-based, or reputation-based solutions to reduce incentive misalignment.
Q: What inspiration and guidance do you think retailers could draw from the Toyota Production System (TPS)?
A: TPS highlights include the power of focusing and executing the operational details; the vital role of continuous process improvement; and the need to involve employees and other firms in the supply chain in improving processes.
Retailers can certainly learn from the auto industry's experience. Like in manufacturing a few decades ago, there is a lot of low-hanging fruit in retail operations today. However, to harvest these fruits retailers will have to
* make problems visible,
* provide their associates with tools for identifying and solving these problems,
* engage their employees and suppliers in their improvement efforts.
Q: Your book notes that typical stores experience staff turnover of 100 percent each year. How do you see successful retailers retain and empower employees?
A: Successful retailers empower their workers to improve processes at the store and also give them the tools and training needed to use the empowerment appropriately. Enlightened retailers also realize the importance of not overloading their store employees with too many tasks. Finally, we do see a number of retailers that are paying careful attention to the hiring, training, and rewarding of store employees.
Our favorite example of a retailer that has used labor successfully is the online retailer Zappos.com. The company pays very careful attention to hiring and training. It reduces turnover by providing employees with a career and seeks to find employees who are looking for one, too. In fact, during the training period, all trainees are offered $2,000 to quit Zappos: Most decline the offer. Zappos's approach to managing its employees results in very satisfied employees, which has translated into very high customer satisfaction and sales growth.
Q: What typical errors do you see retailers make assessing and investing in new technologies? How can retailers overcome these errors?
A: Retailers make several errors concerning new technologies, but they can be surmounted.
Keep an eye on emerging technologies. Technology has the ability to change retailing substantially: The Internet is the most recent example. Hence, retailers—even though they are not usually technology specialists—should pay attention to technology and technological change.
Before adopting the technology, understand its key details. Adopting a technology without understanding its power and limitations usually ends in disaster.
To leverage a technology optimally, pay careful attention to how the technology integrates with the company's operations and business model. When adopting technology, retailers—and many organizations—often fail to make changes to their processes and business model that would be needed to use the technology optimally. That is akin to replacing your car with a helicopter but not making any changes to your lifestyle.
Adopt a "call options" approach. Why? It is usually hard to estimate precisely the benefits that will accrue to a retailer from adopting a new technology: This often leads to retailers being reluctant to invest in the new technology. But by structuring your investment as a "call option," if the technology turns out to meet or exceed expectations, you can make a more substantial investment later. On the other hand, if the technology disappoints, you would have lost only the small investment up front.
Q: What are the main barriers you see for integrating new analytics within retail organizations?
A: In our book we identify many barriers:
* Analytics often require people to change the way they think. This is hard.
* Many non-analytically inclined managers are skeptical of analytical approaches because past claims by some analysts have been exaggerated.
* The retail organization might not have the skills needed to sustain the analytic approach.
* Analytic skills often need organizational change that the retailer is reluctant to make.
* Implementing analytic skills might require collaboration across multiple functions.
* Analytics are simply not "fun" for many people.
Q: How can managers ease the path to adopting the new science of retailing in lieu of making major cultural or structural changes?
A: When it comes to implementing science in retailing, the "missionary" is more important than the scientist. Implementing the science is often a bigger challenge than devising the science itself. Over the years we have seen many successful missionaries implement analytic approaches without the mandate to make large-scale cultural or structural changes.
Here is what some of these missionaries do:
* Recognize that there are multiple users, each of whom has "veto power" on adopting analytics. Involve each of these users and understand and address every one of their concerns.
* Quantify the benefits—and acknowledge the fact that precise quantification is usually impossible! Ideally, "own a benefit" with senior management.
* It's better to win quickly rather than seek to win big!
Book excerpt from The New Science of Retailing: How Analytics Are Transforming the Supply Chain and Improving Performance
By Marshall Fisher and Ananth Raman
How will rocket science retailing be different in the new normal than it was in the past? Equally important, what are the salient aspects of rocket science retailing that need to be reinforced? In the new normal, managers will have to learn to cope not only with mix uncertaintybut also with aggregate demand uncertainty. In addition, managers will have to redouble their effort to improve product availability and operational execution.
Anticipate and React to Fluctuations in Aggregate Demand
With total revenue likely to fluctuate much more than in the past, managers must be ready to take anticipatory action. Consequently, they should improve their ability to forecast aggregate demand. Supply chains during the last few decades have focused on dealing with mix uncertainty. In fact, many of us—rightly for the time we were focused on—assumed that firms could forecast their aggregate demand well but not their mix, and we worked on improving approaches for mix forecasting. Unfortunately, these tools don't translate well to situations with aggregate demand uncertainty.
How well do managers forecast total revenue? You can get a sense of that from looking at investment analysts' forecasts. While it is true that managers have information that analysts lack, analysts' estimates and management's guidance on earning (as opposed to sales) show that the two groups are closely aligned in their thinking. Analysts are often in frequent and close contact with management, even though they're not privy to inside information.
Examining analysts' sales forecasts in 2008 reveals that analysts—and probably managers too—were slow to update their demand forecasts as the economy deteriorated and the financial crisis intensified. […] In March and April 2008, when [Abercrombie & Fitch] was making operational plans (for example, sourcing raw material and planning production), analysts were expecting its annual sales to be roughly $4.2 billion. As late as August 2008, they were still expecting more than $4 billion, and even in October—that is, after Lehman Brothers declared bankruptcy—they were projecting close to $3.8 billion. In other words, they overestimated by roughly $250 million in October, nine months into the year for which they were forecasting. The evidence that we've seen suggests that this performance is typical: analysts and managers have trouble making forecasts in the kind of volatile economy that will probably be part of the new normal.
Pay Attention to Product Availability
Repeated studies in supermarkets and other retailers of fast-moving good have shown that roughly 8 percent to 10 percent of the SKUs in a store are stocked out at any given time. Stockouts are expensive for retailers for two reasons. First, they lead to lost sales. A 2004 study that surveyed over 70,000 consumers in 29 countries found that when they did not find the exact item they wanted, nearly one-third went to another store to buy the product, while less than half bought a substitute.3 Perhaps most significantly, stockouts harm customer goodwill. A study conducted by a multinational consumer-goods maker shows that consumers blame retailers for the stockout a whopping 83 percent of the time, irrespective of who caused it.
Pilot programs at many retailers have shown that stockouts can be reduced with better forecasting, inventory planning, and in-store execution—exactly the techniques described in this book.
Companies often understate sales lost due to stockouts. Our favorite example comes not from retailing but from Hugo Boss Bodywear, a division of Hugo Boss AG. In a test, the bodywear division recently moved 45 of its SKUs from monthly to weekly ordering while leaving the ordering process for 269 SKUs, which served as a control, untouched. The in-stock rate on the 45 weekly SKUs went from 98.24 percent to 99.96 percent. Sales for the 45 SKUs increased by 32 percent, even while sales for the control SKUs fell by 10 percent. Why did sales rise by 32 percent rather than 1.72 percent, the difference between 98.24 percent and 99.96 percent? Though the division had historically achieved high in-stock rates on average, there were periods when in-stock rates on popular styles dropped to roughly 85 percent. Anticipating periods of low availability, retailers often carried an "insurance brand," which they could order more of during the periods when Boss's products were in short supply. By taking the in-stock rate close to 100 percent, Boss reassured retailers that certain SKUs would never be out of stock. Many retailers responded by dropping the insurance brand from their assortments, causing Boss's sales to skyrocket.
Improve Operational Execution
Steve Kaufman, former CEO of Arrow Electronics, a large distributor of electronic components, often says, "A less than perfect strategy with perfect execution will beat a perfect strategy with less with perfect execution every time." The adage is definitely true in retailing.
Execution can take many forms in retailing. In stores, for example, you must greet customers appropriately, keep inventory accurate, and minimize misplaced product. In any economy, these tasks will remain vital to competitiveness. Improving them requires attending to basics like employee training, appropriate staffing, and store layout. Everyone in retailing knows this, but basics are boring, and it's easy to lose sight of them.
The U.S. automotive industry offers a cautionary tale of the perils of not tending to the basics. The Big Three—General Motors, Ford, and Chrysler—declined and Toyota rose primarily because Toyota, with its push for continuous improvement, focused on operational excellence and produced more reliable, economical cars. Not surprisingly, Toyota's market capitalization during the last few years has exceeded by substantial amounts that market capitalizations of Ford, GM, and Chrysler combined—and that was true even before GM filed for bankruptcy and Chrysler was merged with Daimler.
Retailers should learn from the Big Three's mistakes. Toyota couldn't have beaten them without innovative human resource practices that engaged and empowered its line workers to identify ways to improve its production processes. Retailers, too, must recognize that to excel at execution, they must empower their people, including in their stores and distribution centers.
A few years ago, we visited a Toyota factory where we asked the plant manager why Toyota allowed other manufacturers (including competitors) to tour its plants. "Wouldn't they be able to copy the Toyota Production System?" we asked. "Others cannot replicate our performance unless they can replicate what goes on in our people's heads," he said. The more we learned about the Toyota Production System, the more we agreed. Other companies could (and did) easily copy physical attributes of the Toyota Production System, such as andon cords and kanbans. But they couldn't replicate Toyota's approach to people. The challenge with retail execution is similar. Not only do retailers, like manufacturers, have to focus on operational details, they also need to transform their frontline workers into a "community of scientists."
As we discussed earlier in the book, the hardest part of deploying analytics is implementation. It can take a long time and must be done in phases. It's a long journey for most companies. But as the Chinese proverb says, "A journey of a thousand miles begins with a single step." Are you ready to take a step?
[This article was provided with permission from Harvard Business School Working Knowledge.]
