w rP os t S 906D06 op yo WILKINS, A ZURN COMPANY: DEMAND FORECASTING Professors Carol Prahinski and Eric O. Olsen prepared this case solely to provide material for class discussion. The authors do not intend to illustrate either effective or ineffective handling of a managerial situation. The authors may have disguised certain names and other identifying information to protect confidentiality. Ivey Management Services prohibits any form of reproduction, storage or transmittal without its written permission. This material is not covered under authorization from CanCopy or any reproduction rights organization. To order copies or request permission to reproduce materials, contact Ivey Publishing, Ivey Management Services, c/o Richard Ivey School of Business, The University of Western Ontario, London, Ontario, Canada, N6A 3K7; phone (519) 661-3208; fax (519) 661-3882; e-mail cases@ivey.uwo.ca. Copyright © 2005, Ivey Management Services Version: (A) 2006-09-13 tC On Monday, January 10, 2005, as Bernie Barge, the newly promoted inventory manager at the Wilkins plant in Paso Robles, California, prepared for the forecasting meeting scheduled for the following day, he wondered if he could find an easier and possibly more reliable means of forecasting the sales demand. BACKGROUND No Wilkins Regulator Company had built its strength on high-quality products for the plumbing, municipal waterworks, fire production and irrigation customer markets, ranging from water pressure reducing valves and backflow preventers to anti-scald shower valves. The general plumbing customer market represented approximately half of its sales revenue and the irrigation customer market represented approximately a quarter of its sales revenue. Chris Connors, the plant’s general manager and Barge’s supervisor, had targeted the fire protection and municipal waterworks customers as opportunities for growth. Do Zurn Industries acquired Wilkins in 1971. In 1998, Zurn merged with U.S. Industries Bath & Plumbing Products Co., and changed its name to Jacuzzi Brands in 2003. From the most recent Jacuzzi Brands 2004 Annual Report, Barge read: Demand for our products is primarily driven by new home starts, remodeling and construction activity. Accordingly, many external factors affect our business including weather and the impact of the This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 2 broader economy on our end markets. Weather is an important variable for us as it significantly impacts construction. Spring and summer months in the U.S. and Europe represent the main construction season for . . . commercial and industrial markets. As a result, sales in our bath products and plumbing products segments increase significantly in our third and fourth fiscal quarters as compared to the first two quarters of our fiscal year. The autumn and winter months generally impede construction and installation activity. op yo Our plumbing products business is dependent upon commercial and institutional construction activities and is, therefore, affected by macroeconomic factors, such as the unemployment rate and the availability of financing. Despite the cyclical nature of the U.S. commercial and institutional construction market, which experienced declines in revenue of approximately 14 per cent in fiscal 2002, approximately six per cent in fiscal 2003 and a slight rebound in fiscal 2004, sales of our commercial and institutional products have continued to grow at a rate that exceeds that of the industry. We have achieved this growth through favorable pricing, product innovation and targeted marketing programs. tC Connors provided Barge with additional insight into the complexities involved in forecasting demand: No There are lots of uncontrollables. Uncontrollables included the weather, competitors’ product introductions and our own product introduction. Sometimes, we cannibalize our own sales — intentionally — and, sometimes, unintentionally. Other influences on the demand include marketing strategies, such as price promotions and, in the irrigation market segment, an early-buy program that encourages customers to place their orders in the early spring. CURRENT FORECASTING PROCESS: THE FORECAST MASTER Do Each quarter, Connors and Rick Fields, the sales/marketing manager, developed the quarterly demand forecasts for each product family. Barge, in his newly created position, would also participate in the forecast development. Based on their knowledge of industry trends, competitive strategies and sales history, they would estimate the sales for the next five or six quarters. Barge commented: Rather than forecast the total quarterly sales volume for a product family, we forecast the average anticipated sales per week for the quarter for each product family. We have about 25 different This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 3 op yo product families and each product family has what we call a planning bill. To start the process, however, we start with what we call the forecast master. The forecast master is a spreadsheet that lists the average weekly sales history for each product family by quarter and year since 1999. For each product family, we divide the total quarter’s actual sales by 13 weeks per quarter to determine the average weekly sales per quarter. Then, we plug in our expected demand for the next five or six quarters. These numbers represent our best estimate. This information is then used to calculate the average dollars per unit and average gross profit per unit, which is used by our accounting and finance group to develop various budgets. A portion of last quarter’s forecast master is shown in Exhibit 1 for two product families: Pressure Vacuum Breakers (PVBs) and Fire Valves. PVBs were a type of backflow prevention device, which was designed to prevent the reverse flow of water and other substances into the water source. PVBs were used predominantly by the irrigation market segment. tC Fire valves, a type of pressure reducing valve, were designed to reduce or regulate water pressure in residential, commercial and industrial applications. In addition to having just signed on a new customer, Connors anticipated high growth in the fire valve market since Wilkins was introducing a number of new product extensions designed to increase market share. One such product extension was the development of fixed-setting fire valves. Wilkins’ current fire valves had adjustable settings, which were set by the installer. Some regulatory agencies, however, were concerned about improper installation or modification to the valves and were now requiring fixed setting valves to improve safety. No CURRENT FORECASTING PROCESS: THE PLANNING BILL Do Each product family had its own planning bill. Barge described the planning bill: There are five important components to each planning bill. First, the planning bill contains the sales history for each product within the family. We have quarterly sales history that goes back to 1989. If I dig into the old files, I can go back even further. Second, for the last four quarters, the planning bill calculates the average number of units sold within that product family each day within each quarter. For example, for our first fiscal quarter of 2005, which started on October 1, 2004, we sold 48,159 units within the PVB product family [as shown in Exhibit 2]. Since the quarter had 58 days, the planning bill will calculate that we sold a daily average of 830 units. We will also calculate the average daily sales for the last four quarters; for the PVBs, it was 1,205. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 4 Third, the planning bill contains our projection on the average daily sales for that family that we think we will sell in the next 12 months. This number came from the forecast master and is one of the key determinants of the forecast by product. With the PVB product family, for example, we think our sales will have a moderate growth rate predominantly due to industry growth and some problems at one of our competitor’s manufacturing facilities. Last quarter, we forecasted that we would sell an average of 1,400 units each business day in the next 12 months [as shown in the far right column of Exhibit 3]. op yo Fourth, we disaggregate the family forecast into each product based on the per cent of sales of the product family. To do this, we first calculate the proportion of unit sales that each product currently represents within the family. We call this the “raw per cent.” Then, we try to forecast the percent of family sales that the product will represent in the future, which we call the “planning bill per cent.” It can get pretty complicated. If we have new products, we have to factor in the effect that they may have on our existing products. Plus, with new products, we also have to project growth without any historical data. tC The fifth key piece of information in the planning bill is the calculation of the annual sales forecast for each product within the family. We use a couple of key pieces of information: The planning bill per cent is multiplied by 250 days in a year and by the daily sales forecast for the family, which are 1,400 units in this situation. No The sales history for a select group of products is shown in Exhibit 2. The planning bills for the PVBs and fire valves, as of October 2004, which was the beginning of the 2005 fiscal year, are shown in Exhibit 3 and 4, respectively. FORECASTING PERFORMANCE Do When contemplating the forecast accuracy, Barge said, “I don’t have a clue on how well we have been doing. I think we are doing OK at the aggregate level, but we probably have some swings in our accuracy level at the individual product level.” For the first quarter of 2005, Connors and Fields had forecasting sales of 53,560 PVB units and 559 five valve units. According to Exhibit 2, actual sales were 48,159 PVB units and 580 fire valve units. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 IMPLEMENTATION CONCERNS rP os t Page 5 Barge wondered if he could use statistical forecasting methods to ease the forecasting process and perhaps improve the reliability of the sales forecast. If Barge was going to recommend a new forecasting method, he considered how he should gain buy-in from Connors and other managers at the plant. He knew that Connors considered it important to use judgment in developing the sales forecast. For example, if Connors believed that the industry was entering a mild recession, he wanted to make sure the demand forecast reflected the anticipated economic downturn. op yo Barge also wondered how to incorporate the occasional price promotions that were used to sell off excessive finished goods inventory. He knew that if management reduced the price, Wilkins was going to sell more units and be more competitive. tC Barge frequently joked that the fire valves were leading economic indicators. Although he said it jokingly, he wondered if there was some truth in it. Since the product was used in new construction, an increase in product sales would indicate that the construction industry was in an upswing. To help determine the demand forecasts, he wondered if he could use the United States economic information, such as the unemployment rate data (see Exhibit 5), the bank prime loan rates (see Exhibit 6) or the number of new housing starts (see Exhibit 7). Barge knew that less than one per cent of the PVB sales were outside the United States and he didn’t remember any fire valves having been sold outside of the United States. No Finally, he wondered how to forecast new products, such as the new fixed-pressure fire valves. Although he could use the historical sales of the adjustable-pressure fire valves, both he and Connors believed that the new fixed-pressure valves would have dramatic growth, which Barge did not think could be captured by the historical sales data of the older products. As Barge reflected on his preparation for tomorrow’s meeting, he wondered what he should recommend to Connors and how to address any potential implementation concerns. Do The Wilkins series of cases is dedicated to Dr. Michael F. Pohlen, Professor Emeritus of Operations Management, Alfred Lerner College of Business, University of Delaware. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 Exhibit 1 rP os t Page 6 FORECAST MASTER AS OF OCTOBER 7, 2004 WILKINS REGULATOR DIVISION UNIT SALES PER WEEK BY QUARTER FORECAST BEGINNING QI 2005 HISTORY THROUGH Q IV 2004 REVISED OCTOBER 7, 2004 Final FILE: FCSTMASTER FISCAL YEAR 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 FIRE VALVE 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Q II Q III 1788 2097 2116 2352 2721 3029 4120 3748 4008 3523 4092 4449 5786 7480 5115 5532 5921 6824 7184 9451 9341 3167 3123 3374 3968 4531 4231 5983 179,634 191,880 194,136 224,072 245,506 292,465 350,012 36 47 21 36 42 22 43 43 32 45 38 27 26 51 18 34 19 28 25 28 51 13 24 38 43 11 29 51 1,432 1,777 1,608 1,886 1,357 1,371 2,550 No tC PVB Q IV YEAR TOTAL QI op yo PROD LINE Note: Numbers in bold font represent actual weekly unit sales, averaged over the quarter. Do Source: Company files. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. t s o P r Page 7 9B06D006 Exhibit 2 SALES HISTORY FROM THE PLANNING BILLS (as of January 10, 2005) Year Qtr PVB 12-720 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 2003 2003 2004 2004 2004 2004 2005 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2108 2516 2954 2224 1788 2512 3540 3666 3546 5596 5948 4892 4311 6637 7823 6652 6235 o D PVB 34-720 6716 11510 18438 10456 9464 12942 21640 11650 11270 15322 19424 13310 10214 20983 27492 19535 28020 PVB 1-720 10688 21688 40814 19510 9824 21848 48936 24640 12400 23692 47904 26990 13851 31507 66644 17337 11163 N t o Total Fire Valve Z2105 Fire Valve Z3000 Fire Valve Z3000IL Fire Valve Z3004 27512 45798 76968 43858 30580 53198 88704 51590 35372 57840 93388 58906 39382 75219 122868 54996 48159 122 139 54 78 18 224 138 193 124 112 25 37 79 71 68 54 85 77 240 138 275 301 133 175 251 213 150 64 57 108 105 151 114 199 1 0 3 36 -35 0 0 10 2 3 139 3 49 52 2 84 140 20 144 25 59 48 35 15 50 32 46 26 18 23 26 62 44 66 o y p o C PVB 34-420 PVB 1-420 7158 7034 9026 7254 6968 9156 8134 6958 5208 6806 10446 7960 7397 9043 11233 7496 384 842 3050 5736 4414 2536 6740 6454 4676 2948 6424 9666 5754 3609 7049 9676 3976 2357 Fire Valve Z3004IL 50 67 32 49 132 103 42 63 170 40 77 22 26 90 77 86 90 Total 270 590 252 497 464 495 370 567 541 351 331 137 285 344 360 382 580 Note: Data reflects total units sold within the fiscal quarter Source: Wilkins Plant Data This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 Exhibit 3 rP os t Page 8 PLANNING BILL FOR PVBS (as of October 3, 2004) WILKINS DIVISION PLANNING BILL: PVB Revised 10/03/04 cc: Steve, Jim, Cyd, Neal, Terri, Ed, Lisa QI UNIT SALES 09/28/03 01/04/04 Q II UNIT SALES 01/04/04 04/03/04 Q III UNIT SALES 04/04/04 07/03/04 Q IV UNIT SALES 07/04/04 10/02/04 12-720 34-720 1-720 34-420 1-420 4311 10214 13851 7397 3609 6637 20983 31507 9043 7049 7823 27492 66644 11233 9676 TOTAL DAYS UNIT/DAY 39382 64 615 75219 65 1157 122868 64 1920 tC TOTAL DAYS UNIT/DAY RAW % 6652 19535 17337 7496 3976 8.7% 26.7% 44.2% 12.0% 8.3% 10.6% 25.1% 44.1% 12.0% 8.2% 37100 87850 154350 42000 28700 54996 63 873 100.0% 100.0 350000 250 1400 op yo PART NO FORECAST 12-Month DEMAND 1400/DAY PLAN BILL % 292465 256 1142 350000 250 1400 Do No Source: Wilkins Plant Data. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 Exhibit 4 rP os t Page 9 PLANNING BILL FOR FIRE VALVES (as of October 12, 2004) WILKINS DIVISION PLANNING BILL: Fire Valves Revised 10/12/04 cc: Jim, Cyd, Steve, Ed, Terri, Brad, Pete, Lisa QI UNIT SALES 09/28/03 01/04/04 Q II UNIT SALES 01/04/04 04/03/04 Q III UNIT SALES 04/04/04 07/03/04 Q IV UNIT SALES 07/04/04 10/02/04 Z2105 Z3000 Z3000IL Z3004 Z3004IL 79 108 49 23 26 71 105 52 26 90 68 151 2 62 77 TOTAL DAYS UNIT/DAY 285 64 4 344 65 5 360 64 6 tC TOTAL DAYS UNIT/DAY RAW % 54 114 84 44 86 19.8% 34.9% 13.6% 11.3% 20.4% 21.2% 38.8% 11.8% 8.2% 20.0% 540 990 300 210 510 382 63 6 100.0% 100.0 2,550 250 10.2 op yo PART NO FORECAST 12-Month DEMAND 10.2/DAY PLAN BILL % 1371 256 5 2550 250 10.2 Do No Source: Company files. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 10 Exhibit 5 UNEMPLOYMENT RATE IN THE UNITED STATES Unemployment Rate (%) 3.90 4.23 4.40 4.83 5.53 5.70 5.83 5.73 5.87 5.83 6.13 6.13 5.87 5.67 5.57 5.43 5.43 op yo Oct-Dec 2000 Jan-Mar 2001 Apr-Jun 2001 Jul-Sep 2001 Oct-Dec 2001 Jan-Mar 2002 Apr-Jun 2002 Jul-Sep 2002 Oct-Dec 2002 Jan-Mar 2003 Apr-Jun 2003 Jul-Sep 2003 Oct-Dec 2003 Jan-Mar 2004 Apr-Jun 2004 Jul-Sep 2004 Oct-Dec 2004 United States Unemployment Rate Oct-04 Jul-04 Apr-04 Jan-04 Oct-03 Jul-03 Apr-03 Jan-03 Oct-02 Jul-02 Apr-02 Jan-02 Oct-01 Jul-01 Apr-01 Jan-01 Oct-00 No tC 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 Seasonal Adjusted Unemployment Rate Do Note: Data reflects the seasonal adjusted unemployment rate for people 16 years and over. The quarterly data represents the unweighted average of the monthly data. Source: U.S. Department of Labor, Bureau of Labor Statistics. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 11 Exhibit 6 BANK PRIME LOAN RATE IN THE UNITED STATES Bank Prime Loan Rate (%) 9.50 8.62 7.34 6.57 5.16 4.75 4.75 4.75 4.45 4.25 4.24 4.00 4.00 4.00 4.00 4.42 4.94 op yo Oct-Dec 2000 Jan-Mar 2001 Apr-Jun 2001 Jul-Sep 2001 Oct-Dec 2001 Jan-Mar 2002 Apr-Jun 2002 Jul-Sep 2002 Oct-Dec 2002 Jan-Mar 2003 Apr-Jun 2003 Jul-Sep 2003 Oct-Dec 2003 Jan-Mar 2004 Apr-Jun 2004 Jul-Sep 2004 Oct-Dec 2004 Rate of Interest in Money and Capital Markets 9.00 8.00 7.00 No 6.00 tC 10.00 5.00 4.00 Oct-04 Jul-04 Apr-04 Jan-04 Oct-03 Jul-03 Apr-03 Jan-03 Oct-02 Jul-02 Apr-02 Jan-02 Oct-01 Jul-01 Apr-01 Jan-01 Oct-00 3.00 Do Bank Prime Rate Note: Figures are based on the rate posted by majority of top 25 (by assets in domestic offices) insured U.S.-chartered commercial banks. Prime is one of several base rates used by banks to price short-term business loans. Not seasonally adjusted. Quarterly figures are based on the unweighted average of the monthly rates, which include each calendar day in the month. Source: Federal Reserve Statistical Release, H.15, Selected Interest Rates. This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. 9B06D006 rP os t Page 12 Exhibit 7 NEW PRIVATELY OWNED HOUSING UNITS STARTED IN THE UNITED STATES (in 000s) Multi-Unit Housing Starts 82 74 87 88 81 76 89 98 84 71 84 99 95 80 84 92 90 Total Housing Starts 357 348 461 429 366 369 475 459 403 375 490 511 472 425 540 532 460 op yo Oct-Dec 2000 Jan-Mar 2001 Apr-Jun 2001 Jul-Sep 2001 Oct-Dec 2001 Jan-Mar 2002 Apr-Jun 2002 Jul-Sep 2002 Oct-Dec 2002 Jan-Mar 2003 Apr-Jun 2003 Jul-Sep 2003 Oct-Dec 2003 Jan-Mar 2004 Apr-Jun 2004 Jul-Sep 2004 Oct-Dec 2004 Single-Unit Housing Starts 275 274 374 341 285 293 386 361 319 304 406 412 377 345 456 440 370 tC New Privately Owned Housing United Started 600 500 400 300 No 200 100 Multi-Units Oct-04 Jul-04 Apr-04 Jan-04 Oct-03 Jul-03 Apr-03 Jan-03 Oct-02 Jul-02 Apr-02 Jan-02 Jul-01 Apr-01 Jan-01 Oct-01 Single Units Total Do Oct-00 0 Source: U.S. Census Bureau: Manufacturing, Mining and Construction Statistics This document is authorized for use only by SAED SALHIEH until June 2011. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or 617.783.7860. Page 2 of 19 2009 1. 2. 3. How accurate was the demand for the first quarter of 2005? What is the current demand forecasting method? Who uses the demand forecast? What are the consequences, if any, if the forecast was inaccurate? Create a demand forecast for the PVB product family for the next three quarters. How does your forecast compare to Wilkin's demand forecast? As Bernie Barge, what would you recommend to management and why? How should Barge convince management to follow his recommendations? Develop an action plan. 4.
Purchase answer to see full attachment