Unless otherwise indicated, all quotations from retailers are from discussions, interviews, or talks given at a class or conference with the authors.

Source: S&P’s Industrial Annual Database accessed through Wharton Research Data Services (WRDS).

See Marshall L. Fisher, Ananth Raman, and Anna Sheen McClelland, “Rocket Science Retailing Is Almost Here: Are You Ready?” Harvard Business Review, July–August 2000, 115–124, for a more extensive description of this study.

For a discussion of David Berman’s business model and investment approach, see Ananth Raman, Vishal Gaur, and Saravanan Kesavan, “David Berman,” Case 605-081 (Boston: Harvard Business School, 2005).

Nardelli had worked at General Electric (GE) before taking over as CEO of The Home Depot.

A periodic report where Berman discusses his thoughts on retail, focusing on inventories.

Raman, Gaur, and Kesavan, “David Berman.”

Ibid.

Saravanan Kesavan, Vishal Gaur, and Ananth Raman, “Incorporating Price and Inventory Endogeneity in Firm-Level Sales Forecasting” (working paper, Harvard Business School, Boston, 2009).

“No Mood For Shopping—Part II,” Barron’s, June 12, 2006.

See Richard Lai, “Inventory Signals” (doctoral candidate research paper, Harvard Business School, 2006.

Inventory turns are calculated by dividing the retailer’s annual cost of sales by the average level of inventory at a firm. The calculation for an analogous term, days of inventory, is similar; to derive days of inventory, the firm’s average inventory level is multiplied by 365 (i.e., the number of days in a year) and then divided by the cost of sales. Some practitioners, including David Berman, have argued to us that days of inventory is more intuitive and hence, more easily accessible to people. We agree but continued using inventory turns because a number of our technical results (e.g., the regression analysis reported later in the chapter) were easier to explain using inventory turns.

A detailed discussion of the results in this section can be found in Vishal Gaur, Marshall Fisher, and Ananth Raman, “An Econometric Analysis of Inventory Turnover Performance in Retail Services,” Management Science 51, no. 2 (2005): 181–194.

Using the pooled model, adjusted inventory turns can be calculated from the following equation: log (adjusted inventory turns)_it = log(IT)_it + 0.2431 log MU_it–0.2502 log CI_it–0.143 log SS_it.

Gross margin is calculated as the difference between the sale price and the cost of goods sold, divided by the sale price. It is usually expressed as a percentage.

A formal argument for this logic can be found in the newsboy model or the newsvendor model, which is usually described in most texts on inventory management ; e.g., Steven Nahmias, Production and Operations Analysis, 5th ed. (Boston: McGraw-Hill Irwin, 2005).

Gaur, Fisher, and Raman, “An Econometric Analysis of Inventory Turnover,” calculate capital intensity as the ratio of the gross fixed assets (i.e., the noninventory assets) to the total assets at a retailer (which comprises inventory and gross fixed assets).

We conducted our analysis with multiple forecasts based on past sales data.

We were introduced to this example by Vishal Gaur.

Vishal Gaur, presentation at the Consortium for Operational Excellence in Retailing, Philadelphia, PA, Wharton School, University of Pennsylvania, June 2006.

Vishal Gaur, Nikolay Osadchiy, and Sridhar Seshadri, “Sales Forecasting with Financial Indicators and Experts’ Input,” Johnson School Research Paper Series No. 06-09, October 23, 2008.

Our discussion of current practice is based on Michael Levy and Barton Weitz, Retailing Management (New York: Irwin-McGraw Hill, 2007) and conversations with several retail executives, including Kevin Freeland, chief operating officer, Advance Auto Parts; Herb Kleinberger, principal, ARC Business Advisors; and Rob Price, chief marketing officer, CVS.

Sometimes the term merchandise planning is used for what we are calling strategic assortment planning.

This discussion is based on Zeynep Ton and Ananth Raman, “Borders Group, Inc.,” Case 9-601-037 (Boston: Harvard Business School, 2003).

Subsequent to BJ’s, Mr. Halpin was CEO of CompUSA and is currently founder and CEO of River Bend Inc.

See Jena McGregor, “At Best Buy, Marketing Goes Micro,” BusinessWeek, May 15, 2008; Vanessa O’Connell, “Reversing Field, Macy’s Goes Local,” Wall Street Journal, April 21, 2008; Ann Zimmerman, “To Boost Sales, Wal-Mart Drops One-Size-Fits-All Approach,” Wall Street Journal Online, September 7, 2006; and Ann Zimmerman, “Home Depot Learns to Go Local,” Wall Street Journal, October 7, 2008, for descriptions of efforts by Best Buy, Macy’s, Wal-Mart, and Home Depot to localize their assortments.

The material on operational assortment optimization is based on Marshall Fisher and Ramnath Vaidyanathan, “Retail Assortment Optimization: An Attribute Based Approach” (white paper, Wharton School Operations and Information Management, Philadelphia, PA, September 2008; revised May 2009).

The total demand estimate column of table 2-6 is an estimate calculated as total sales divided by percent share of demand captured, as described above. The total demand column gives the total of the four brand-size estimates. The two differ for any flavor for which demand estimates were replaced by actual sales.

This estimation approach, while relatively simple, has some weaknesses. For one thing, the chain offered all four brand-sizes for two flavors, which made the calculations easier. In most cases, you can’t count on this happening. Also, by using only two flavors to estimate shares, you’re throwing away lots of data that could be used to improve the accuracy of the estimates. In the butter, cheese, raspberry, vanilla, and honey flavors, this retailer offered both Yummy Cakes and Tiny Tina in the single-serving size, so these sales are undistorted by substitution and could be used with sales in the chocolate and cinnamon categories to more accurately estimate the relative demand shares of Yummy Cakes and Tiny Tina in the single-serving size. All sales data is somewhat influenced by random events (weather, the ever-fluctuating economic news, etc.), so the more data you can bring to bear on estimating a value, the more accurate your estimate is likely to be.

For these reasons, in practice, you’ll employ a statistical method called maximum likelihood estimation, which uses all of the sales data to find brand-size shares and substitution frequency estimates. You’ll apply this technique to sales data for each store to capture store-specific differences in customer preferences. To estimate price on SKUs not currently carried by a retailer, you regress the prices of existing SKUs against their attributes to obtain an attribute-based pricing formula. For additional details, see Fisher and Vaidyanathan, “Retail Assortment Optimization.”

Canadian Tire Corporation Ltd., Annual Report 2007.

The idea of using stockouts as an opportunity to estimate substitution was suggested to us by Kevin Freeland.

The methodology described here was developed in collaboration with one of our former doctoral students, Kumar Rajaram, now a professor at UCLA’s Anderson School. Additional details are reported in Marshall Fisher, Kumar Rajaram, and Ananth Raman, “Optimizing Inventory Replenishment of Retail Fashion Products,” Manufacturing & Service Operations Management 3, no. 3 (2001): 230–241.

The committee forecasting process we’ll describe here, and its use in estimating the standard deviation of forecasts, was first developed in a project we conducted, together with Janice Hammond of Harvard Business School, at Sport Obermeyer, a fashion skiwear firm. The idea was suggested by Wally Obermeyer when he was president of Sport Obermeyer and working with us on this project. For more details see Marshall Fisher, Janice Hammond, Walter Obermeyer, and Ananth Raman, “Making Supply Meet Demand in an Uncertain World,” Harvard Business Review, May–June 1994.

In using this approach, we are ignoring for simplicity the fact that the 55 percent value also varies from season to season. This approach makes sense if this variation is small, which it is. We used the relationship that the probability that demand in the first fourteen weeks is less than or equal to X equals the probability that total season demand is less than or equal to X / 0.55¹ to first find a continuous distribution, and then converted this to a discrete distribution using the approach described in table A-1.

The approach to testing described here was developed in collaboration with one of our former doctoral students, Kumar Rajaram, now a professor at UCLA’s Anderson School. Additional details are reported in Marshall Fisher and Kumar Rajaram, “Accurate Testing of Retail Merchandise: Methodology and Application,” Marketing Science 19, no. 8 (2000): 266–278.

Freeland is currently chief operating officer, Advance Auto Parts.

To determine that 2.5 was the price elasticity that best fit the data, we made price elasticity a parameter in a cell of an Excel model containing the historical data, and then computed the forecast error measure mean absolute deviation (MAD) by, for each historical data point, finding the absolute difference between predicted and actual lift and summing across all points. Then we used the Excel function Solver to find the elasticity value that minimized MAD, which turned out to be 2.5. See Stephen Smith and Dale Achabal, “Clearance Pricing and Inventory Policies for Retail Chains,” Management Science, 1998, for additional details on estimating the sales impact of markdowns.

Facing severe competition, Zany Brainy discontinued operation in 2001. Schlesinger is now cofounder and president of Five Below, a leading extreme value retailer to the teen market and beyond.

See Marshall L. Fisher, Ananth Raman, and Anna Sheen McClelland, “Rocket Science Retailing Is Almost Here: Are You Ready?” Harvard Business Review, July–August 2000, 115–124.

This discussion is based on Marshall Fisher, “National Bicycle Industrial Co” (unpublished case), the Teaching Note for this case, and “Japan’s New Personalized Production,” Fortune, October 22, 1990.

A hansha (for “sales company”) is a distributor more or less peculiar to Japan that is incorporated as an independent company but that is completely captive to a particular manufacturer. National Bicycle used ten hansha that had been established by Matsushita Electric and were run by ex-managers of Matsushita. The hansha carried only Panasonic bicycles that they bought from National Bicycle and resold to retailers in exclusive territories.

A servomotor has an output shaft that can be set to a specific angular position based on a signal from a computer. Servos are used in a wide range of applications, from remote-controlled planes to robots.

L. Kopczak and H. Lee, “Hewlett-Packard Company DeskJet Printer Supply Chain (A),” Case GS3A (Stanford, CA: Stanford Graduate School of Business, May 2001).

Dana Canedy, “McDonald’s Burger War Salvo: Is ‘Made for You’ the Way Folks Want to Have It?” New York Times, June 20, 1998.

For additional details, see Marshall Fisher et al., “Making Supply Meet Demand in an Uncertain World,” Harvard Business Review, May–June 1994.

G. N. Georgano, Cars: Early and Vintage, 1886–1930 (London: Grange-Universal, 1985).

One might also argue that labor costs have become a smaller portion of total costs and hence maximizing labor efficiency is less crucial.

This section is based in part on the cases “Supply Chain Management at World, Ltd.,” by Anna Sheen McClelland, Ananth Raman, and Marshall Fisher, Case 9-601-072 (Boston: Harvard Business School, 2001) and “Zara: IT for Fashion,” by Andrew McAfee, Vincent Dessain, and Anders Sioman, Case 9-604-081 (Boston: Harvard Business School, 2007). The facts about World are accurate as of the time of the case.

For additional details on the Obermeyer forecast process, see Marshall Fisher, Janice Hammond, Walter Obermeyer, and Ananth Raman, “Making Supply Meet Demand in an Uncertain World,” Harvard Business Review, May–June 1994.

Freeland is currently chief operating officer, Advance Auto Parts.

The term rocket science retailing is discussed in the Introduction and is used equivalently in this book with scientific retailing. See also Marshall L. Fisher, Ananth Raman, and Anna Sheen McClelland, “Rocket Science Retailing Is Almost Here: Are You Ready?” Harvard Business Review, July–August 2000, 115–124, for a discussion of this concept.

For further details, see V. G. Narayanan and Ananth Raman, “Hamptonshire Express,” Case 1-698-053 (Boston: Harvard Business School, 2002).

Optimal stocking values can be derived from the newsvendor model. A derivation of the newsvendor model can be found in most textbooks on inventory theory—e.g., Steven Nahmias, Production and Operations Analysis, 5th ed. (Boston: McGraw-Hill Irwin, 2005).

Ideas for this section are taken from V. G. Narayanan and Ananth Raman, “Aligning Incentives for Supply Chain Efficiency,” Note 600-110 (Boston: Harvard Business School, April 10, 2000).

See, for example, Michael C. Jensen and Kevin J. Murphy, “CEO Incentives—It’s Not How Much You Pay, But How,” Harvard Business Review, May–June 1990.

Details of the analysis reported in this section can be found in Nicole DeHoratius and Ananth Raman, “Store Manager Incentive Design and Retail Performance: An Exploratory Investigation,” Manufacturing & Service Operations Management 9, no. 4 (2007). Shrink refers to inventory reductions usually associated with theft or merchandise lost from the store.

Zeynep Ton and James L. McKenney, “Campbell Soup Company: A Leader in Continuous Replenishment Innovations,” Case 608-141 (Boston: Harvard Business School, 2008); and Janice H. Hammond, “Barilla SpA (A-D),” Case 694-046 (Boston: Harvard Business School, 2007).

V. G. Narayanan and Lisa Brem, “Owens and Minor, Inc. (A),” Case 100-055 (Boston: Harvard Business School, 2008).

Victor Fung and Joan Magretta, “Fast, Global, and Entrepreneurial: Supply Chain Management, Hong Kong Style: An Interview with Victor Fung,” Harvard Business Review, May 1998.

See Jeffrey Liker and Thomas Choi, “Building Deep Supplier Relations,” Harvard Business Review, December 2004.

V. G. Narayanan and Ananth Raman, “Aligning Incentives in Supply Chains,” Harvard Business Review, November 2004.

See “10-Foot Rule,” Wal-Mart Stores, http://walmartstores.com/AboutUs/285.aspx.

Results for specific retailers are reported in Nicole DeHoratius and Ananth Raman, “Inventory Record Inaccuracy: An Empirical Analysis,” Management Science 54, no. 4 (2008): 627–641; Marshall Fisher, Jayanth Krishnan, and Serguei Netessine, “Retail Store Execution: An Empirical Study” (working paper, Wharton School, Operations and Information Management, Philadelphia, PA, January 2007); Marshall Fisher, Jayanth Krishnan, and Serguei Netessine, “A Cross Sectional Study of Retail Store Performance” (working paper, Wharton School, Operations and Information Management, Philadelphia, PA, October 2007); Zeynep Ton and Ananth Raman, “Cross Sectional Analysis of Phantom Products at Retail Stores” (working paper, 2000); Zeynep Ton and Ananth Raman, “The Effect of Product Variety and Inventory Levels on Retail Sales: A Longitudinal Study,” Production and Operations Management Journal (forthcoming) ; Zeynep Ton, “The Effect of Labor on Profitability: The Role of Quality” (Harvard Business School working paper, 2009); and Zeynep Ton and Robert S. Huckman, “Managing the Impact of Employee Turnover on Performance: The Role of Process Conformance,” Organization Science 19, no. 1 (2008): 56–68.

The results described for this retailer are based on Fisher, Krishnan, and Netessine, “Retail Store Execution.”

Fisher, Krishnan, and Netessine, “Retail Store Execution,” describes two statistical analyses done to determine the most important driver of sales: one using a technique called stepwise regression that adds explanatory variables sequentially in order of importance, and the other using variable normalization so that a variable’s coefficient represents its importance in explaining sales.

The exact methodology used to measure the impact of store labor on sales is described in detail in Fisher, Krishnan, and Netessine, “Retail Store Execution”; but briefly, we had over six thousand observations in our data on store–month sales, planned payroll, and actual payroll, and we were able to analyze this data to measure the impact of payroll on sales.

A detailed description of how store-specific payroll sales lifts were computed is given in Fisher, Krishnan, and Netessine, “Retail Store Execution.” A store that had high sales variation that correlated with a relatively low payroll variation would have a high sales lift, whereas a store that had high payroll variation but little sales variation would have a low sales lift.

Obviously, there will be variation in sales and payroll due to seasonality, but the numbers in figure 6-5 have been deseasonalized.

K. Maher, “Wal-Mart Seeks New Flexibility in Worker Shifts,” Wall Street Journal Online, January 3, 2007.

This discussion is based on conversations with Kevin Freeland, chief operating officer, Advance Auto Parts, and former senior vice president of inventory management, Best Buy.

As reported in “Short-Circuited: Cutting Jobs as Corporate Strategy,” Knowledge@Wharton, April 4, 2007.

May Wong, “Best Buy Service Trumps Circuit City,” Associated Press, April 8, 2007.

“Best Buy Reports 18.5% Increase in Profit,” New York Times, April 5, 2007.

RTT News Global Financial Newswires, “Circuit City Slips to Loss in Q2: Sees Continued Weakness in Q3,” September 20, 2007.

Gaffney is now executive vice president of operations at AAA Northern California, Nevada, and Utah.

The discussion on Staples is based on an interview with Paul Gaffney.

Ton, “The Effect of Labor on Profitability.”

Ton and Huckman, “Managing the Impact of Employee Turnover.”

Nicole DeHoratius and Ananth Raman, “Building on Foundations of Sand?” ECR Journal 3, no. 1 (Spring 2003); Ananth Raman, Nicole DeHoratius, and Zeynep Ton, “Execution: The Missing Link in Retail Operations,” California Management Review 43, no. 3 (Spring 2001); Ananth Raman, Nicole DeHoratius, and Zeynep Ton, “The Achilles Heel of Supply Chain Management,” Harvard Business Review, May 2001, 136–152; DeHoratius and Raman, “Inventory Record Inaccuracy”; and Ton and Raman, “Cross Sectional Analysis of Phantom Products.” For additional information on the status of in-stocks in the grocery industry, see Daniel Corsten and Thomas Gruen, “Stock-outs Cause Walkouts,” Harvard Business Review, May 2004; and Daniel Corsten and Thomas Gruen, “Desperately Seeking Shelf Availability: An Examination of the Extent, the Causes, and the Efforts to Address Retail Out-of-Stocks,” International Journal of Retail & Distribution Management 31, no. 12 (2003): 605–617.

For additional details, see Ton and Raman, “Cross Sectional Analysis of Phantom Products”; and Ton and Raman, “The Effect of Product Variety and Inventory Levels.”

This section is based in part on Marshall Fisher, “To You It’s a Store; To Me It’s a Factory,” ECR Journal—International Commerce Review 4, no. 2 (Winter 2004).

See K. Mishina, “Toyota Motor Manufacturing, U.S.A., Inc.,” Case 0-693-019 (Boston: Harvard Business School, 1995) for an excellent description of the Toyota Production System.

See W. A. Shewhart, Economic Control of Quality of Manufactured Product (New York: D. Van Nostrand Company, Inc., 1931); and W. Edwards Deming, Out of the Crisis (Cambridge, MA: MIT Press, 1986).

DeHoratius and Raman, “Inventory Record Inaccuracy”; and Ton and Raman, “The Effect of Product Variety and Inventory Levels.”

DiRomualdo is currently founder, chairman, and chief executive officer of Naples Ventures, LLC.

See DeHoratius and Raman, “Inventory Record Inaccuracy,” for additional details.

See http://www.toyotageorgetown.com/history.asp.

See W. J. Salmon, “Retailing in the Age of Execution,” Journal of Retailing 65, no. 3 (1989): 368–378, for an excellent overview of the importance of execution in retailing.

The term rocket science retailing is discussed in the Introduction and is used equivalently in this book with scientific retailing. See also Marshall L. Fisher, Ananth Raman, and Anna Sheen McClelland, “Rocket Science Retailing Is Almost Here: Are You Ready?” Harvard Business Review, July–August 2000, 115–124, for a discussion of this concept.

This section draws from Edmund W. Schuster, Stuart J. Allen, and David L. Brock, Global RFID: The Value of the EPCglobal Network for Supply Chain Management (Berlin: Springer-Verlag, 2007).

D. Corsten and T. Gruen, “Desperately Seeking Shelf Availability: An Examination of the Extent, the Causes, and the Efforts to Address Retail Out-of-Stocks,” International Journal of Retail & Distribution Management 31, no. 12 (2003): 605–617.

Zeynep Ton and Ananth Raman, “Borders Group, Inc.,” Case 9-601-037 (Boston: Harvard Business School, 2007).

Zeynep Ton, Vincent Dessain, and Monika Stachowiak-Joulain, “RFID at the METRO Group,” Case 606-053 (Boston: Harvard Business School, 2005).

Brian Harris and James Tenser, “Retail Execution: The Buck Starts Here,” Progressive Grocer, May 1, 2008.

Bill C. Hardgrave, Matthew Waller, and Robert Miller, “Does RFID Reduce Out-of-Stocks? A Preliminary Analysis” (Fayetteville, AK: Information Technology Research Institute, Sam M. Walton College of Business, University of Arkansas), http://itrc.uark.edu.

As an example, note that METRO (from the “RFID at the METRO Group” case) reasoned that reducing out-of-stocks by 2% would cause sales to increase by 0.5%.

Based on 23% gross margins and annual sales of $344 billion.

Robert Hayes and Ramachandran Jaikumar, “Manufacturing’s Crisis: New Technologies, Obsolete Organizations,” Harvard Business Review, September 1988.

Ibid.

Ibid.

Advertisement for International Commercial Truck, circa 1910, on display in Maine’s Owl’s Head Transportation Museum.

Hau Lee, “Peering Through a Glass Darkly,” ECR Journal: International Commerce Review, Spring 2007, articulated this phenomenon first.

Alan MacCormack, “Managing Innovation in an Uncertain World: Module 1: Innovation and Uncertainty,” Module Note 5-606-125 (Boston: Harvard Business School, 2006).

Graham R. Mitchell and William F. Hamilton, “Managing R&D as a Strategic Option,” Research-Technology Management (May–June 1988), reprinted March–April 2007.

Zeynep Ton and Ananth Raman, “Borders Group, Inc.”

The term rocket science retailing is discussed in the Introduction and is used equivalently in this book with scientific retailing. See also Marshall L. Fisher, Ananth Raman, and Anna Sheen McClelland, “Rocket Science Retailing Is Almost Here: Are You Ready?” Harvard Business Review, July–August 2000, 115–124, for a discussion of this concept.

Jim Shepherd, AMR Research.

Freeland is currently chief operating officer, Advance Auto Parts.

See Ananth Raman and Colin Welch, “Merchandising at Nine West Retail Stores,” Case 698-098 (Boston: Harvard Business School, 1998).

Kaufman joined Arrow as corporate EVP and president of Electronics Distribution Division in August 1982, was named corporate president in May 1985, named CEO in September 1986, named chairman in May 1992, stepped down as CEO July 2000 (remaining chairman), was interim CEO June 1–September 15, 2002, and retired fully from the company on September 15, 2002.

From Martin Walker, “The New Normal, Wilson Quarterly 33, no. 3 (Summer 2009).

Martin Walker, “The New Normal,” Wilson Quarterly 33, no. 3 (Summer 2009): 63–66.

Daniel Corsten and Thomas Gruen, “Stock-outs Cause Walkouts,” Harvard Business Review, May 2004.

Cite Hugo Boss case or paper.

This instruction is for Microsoft Excel 2007. If you have a different version, you may need to consult the help command for how to construct a data table.

We created tables 3-6 and A-1 using demand values ranging from 0 to 222, but only show demand from 70 to 121 in table 3-6, since other demand values have negligible probability.