gp721 2014.pdf

MIS AUTOMATION FOR ACCURATE SALES FORECASTING, SALES ANALYSIS AND STOCK PLANNING Graduation Project Submitted in Partial Fulfillment of the Requirements for the degree of Bachelor of Fashion Technology (B.F.TECH)

BY ANKIT GHOSH M/BFT/10/05 Under the guidance of Mr. Ranjan Kumar Saha Department of fashion technology NIFT, Mumbai

NATIONAL INSTITUTE OF FASHION TECHNOLOGY, MUMBAI (2010-14)

Graduation Project | BFT NIFT MUMBAI 2010-2014

1

ACKNOWLEDGEMENT: First and foremost I would like to thanks Pantaloons Fashion Retail Limited which gave me an opportunity to work with their organization. Further, I would like to thank my college mentor Mr. Ranjan Kumar Saha and my company mentor Mr Amitesh Soni who guided me throughout my project. I would also like to thank Mr Joshua Abraham – Consultant A.T.Kearney, Mr Anand Kumaran – MIS Support Team and Mr Sourabh Tiwari – IT Head to whom I was associated during my Graduation Project. I also express my gratitude towards Mr Amir Chavar, Ms Garima Dubey, Mr Neeraj Jagga, and Ms Nisha Varma members of the Planning Team who helped me a lot during my project.


2

ABSTRACT: BACKGROUND: With a larger uncertainty and a more rapid change in today’s business environment, a heavier role to play lies within predicting future sales, also known as sales forecasting. Although prediction becomes more important in order to not lose market shares, not all companies regard the sales forecasting process as a key function within their organization. RESEARCH ISSUE AND OBJECTIVE: Sales forecasting is a common practice in retail industry but little is known about what methods and techniques are used and what the attitude towards the sales forecasting management are. Since sales forecasting works as an important information input to organizational planning, I will empirically analyze and explore the attitudes towards sales forecasting techniques. METHOD: In order to explore and analyze the attitudes towards the sales forecasting process with Pantaloons, various forecasting techniques are used in order to build logic for the MIS which would give the maximum accuracy in sales forecasting. FINDINGS AND CONCLUSIONS: Usually sales forecasting is done either on the intuition of the buyers along with the planning team by considering the current trend in the market or by considering the sales data for the past few years in order to know a particular trend in the sales values across the years, but these practices do not give an accurate sales forecast because today fashion keeps on changing every moment. A best selling merchandise of the last year may not be the best seller this year if the product life cycle fades. Hence forecasting for a particular season should be done on the basis of the sales trends for the first few weeks for that particular season. But the limitation is that for the first few weeks of that season the forecasting is on the basis of the past years data.


3

TABLE OF CONTENTS: 1)

INTRODUCTION ……………………………………………………………………………………………………….5 1.1) PRIMARY OBJECTIVE …………………………………………………………………………………………. 6 1.2) SECONDARY OBJECTIVE …………………………………………………………………………………..…6 1.3) NEED OF THE PROJECT ……………………………………………………………………………………….6 1.4) A COMPREHENSIVE OVERVIEW OF SALES PLANNING………….………………………………6 1.5) FASHION SELLING PERIOD ………………………………………………………………………………….7 1.6) RETAIL PLANNING PROCESS ……………………………………………………………………………….7 1.7) MERCHANDISE PLANNING ………………………………………………………………………………… 7 1.8) MERCHANDISE MANAGEMENT AND SALES FORECAST ……………………………………… 8 1.9) MERCHANDISE PLAN ………………………………………………………………………………………….9 1.10) PURPOSE OF A MERCHANDISE PLAN ………………………………………………………………… 9 1.11) PLANNING STOCK AND INVENTORY CONTROL ………………………………………………… 10

2)

LITERATURE REVIEW ………………………………………………………………………………………………11 2.1) FORECASTING TECHNIQUE ……………………………………………………………………………… 12 2.2) FORECASTING AS PART OF MANAGEMENT PROCESS ……………………………………….13 2.3) PURPOSE OF FORECASTING ……………………………………………………………………………..14 2.4) APPROACHES AND METHOD OF FORECASTING ………………………………………………..14 2.5) ACCURACY IN FORECASTING …………………………………………………………………………… 16 2.6) INFORMATION SEARCH IN MERCHANDISE PLANNING ………………………………………16

3)

RESEARCH METHODOLOGY …………………………………………………………………………………….17 3.1) MANUAL SALES FORECASTING ………………………………………………………………………….18 3.2) DATA COLLECTION …………………………………………………………………………………………….18 3.3) PLANNING SALES GOALS ……………………………………………………………………………………19 3.4) BEGINNING OF MONTH INVENTORY ………………………………………………………………….23 3.5) END OF MONTH INVENTORY ……………………………………………………………………………..23

4)

IMPLEMENTATION AND DATA ANALYSIS ……………………………………………………………….. 25 4.1) PHASE I OF MIS AUTOMATION (BUILDING A SALES FORECASTING TECHNIQUE FOR MIS) ………………………………………………………………………………………..26 4.2) FORECASTING TECHNIQUES ……………………………………………………………………………… 26 4.3) TIME SERIES-PLOT ……………………………………………………………………………………………..27 Graduation Project | BFT NIFT MUMBAI 2010-2014

4

4.4) 4.5) 4.6) 4.7)

ANALYSIS OF THE ABOVE TIME SERIES PLOT ……………………………………………………...30 SELECTING A FORECASTING METHOD ………………………………………………………………..31 FORECAST ACCURACY …………………………………………………………………………………………31 NAÏVE METHOD OF SALES FORECASTING ……………………………………………………………31

4.8) 4.9) 4.10) 4.11) 4.12)

AVERAGE SALES METHOD ………………………………………………………………………………..34 MOVING AVERAGES AND EXPONENTIAL SMOOTHNING ………………………………… 35 MOVING AVERAGES ………………………………………………………………………………………. 36 EXPONENTIAL SMOOTHNING ………………………………………………………………………… 38 COMPARISON OF VARIOUS FORECASTING TECHNIQUES TO FIND OUT THE BEST LOGIC FOR THE MIS ………………………………………………………………. .41 TREND PATTERN OF SALES FORECASTING………………………………………………………. 41 LINEAR TREND REGRESSION FOR MEN’s CATEGORY ………………………………………. 41 LINEAR REGRESSION FOR SALES FORECASTING FOR NON APPS ……………………...44 LINEAR REGRESSION FOR SALES FORECASTING FOR WOMEN WESTERN ………… 46 LINEAR REGRESSION FOR SALES FORECASTING FOR KIDS CATEGORY …………….. 48 PHASE II OF THE MIS AUTOMATION ………………………………………………………………. 50 DESIGNING THE LAYOUT OF THE MIS SALES REPORT …………………………………….. 52 BUILDING THE LOGICS FOR THE CALCULATED FIELDS IN THE MIS REPORT ……………………………………………………………………………………….53 COLOR CODE LOGIC ………………………………………………………………………………………. 54 MIS SALES REPORT …………………………………………………………………………………………55 PHASE III OF THE MIS AUTOMATION ……………………………………………………………..57 DESIGNING THE LAYOUT FOR PHASE III ………………………………………………………….67 LOGICS FOR MIS AUTOMATION PHASE III ………………………………………………………68

4.13) 4.14) 4.15) 4.16) 4.17) 4.18) 4.19) 4.20) 4.21) 4.22) 4.24) 4.25) 4.26)

5)

RESULTS …………………………………………………………………………………………………………….…69 5.1) RESULT FROM THE PHASE I OF MIS AUTOMATION …………………………………………70 5.2) MEN’s CATEGORY …………………………………………………………………………………………..70 5.3) Non Apps CATEGORY ……………………………………………………………………………………..70 5.4) KIDS CATEGORY ……………………………………………………………………………………………..71 5.5) WOMEN WESTERN CATEGORY ……………………………………………………………………….72 5.6) WOMEN ETHNIC CATEGORY …………………………………………………………………………..73

6) CONCLUSION …………………………………………………………………………………………………………….74 6.1) MIS AUTOMIZATION FOR SALES FORECAST CONCLUSION ………………………………75

7) REFRENCES ……………………………………………………………………………………………………………….76 7.1) BOOKS ……………………………………………………………………………………………………………77 7.2) ARTICLES AND JOURNALS ……………………………………………………………………………….77 7.3) WEBSITES ……………………………………………………………………………………………………….77 Graduation Project | BFT NIFT MUMBAI 2010-2014

5

CHAPTER I

INTRODUCTION


6

1) INTRODUCTION 1.1) PRIMARY OBJECTIVE: The primary objective of my project is to provide logics for the atomization of current sales report, the budget planning through sales forecasting using the best mathematical model in order to reduce the forecast errors as well as the stock analysis for the upcoming weeks in order to meet the customer demand. 1.2) SECONDARY OBJECTIVE:  To find out the forecast errors in case of the initial budget planning as well as the forecast logics which are built for the MIS Automation and see how much percentage fluctuations are there in both the budget plan.  Creating a manual report in the same format as that of the Automated Report for both the weekly sales report as well as the sales forecasting report and validating it with the MIS reports to find for the errors and accordingly correcting it to make the MIS system stabilize. 1.3) NEED OF THE PROJECT: As the world continues to develop into a more complex environment, higher demand has grown for trendy products with a following shorter life cycle. Today there is a concept called as fast fashion which means that today’s fashion garments are so cheap to produce that they are almost seen to be disposable. To be able to foresee trends, seasonality and what customers truly demand, increases the odds for a business to show good financial results. One way of decreasing the role of chance, in dealing with this environment, is to use an accurate sales forecast. A forecast can be seen as scientific best guess for a company’s future demand. The argument for this is, with an accurate forecast of future sales a large benefit in especially the purchasing, the production and logistic planning can be gained. The whole system of sales forecasting is being automated for which various logics are required in order to get the best results. 1.4)

A COMPREHENSIVE OVERVIEW OF SALES PLANNING (BACKGROUND STUDY)

Realistic assortment planning for a particular market or product has become a difficult task in today’s consumer environment. Market competition has increased, consumers want more product variety, and consumer needs from a product have become complex and various. Assortments planning for fashion merchandise are more complex as compared to basic merchandise and require sophisticated analysis of fashion and color trends. Hence to reduce Graduation Project | BFT NIFT MUMBAI 2010-2014

7

the uncertainty of merchandising decisions, retail planners search for information from various sources:  Past Sales History  Their own experience  Competitors sales situation Retail Planners most difficult job to meet the consumer demand is the determination of the Stock Keeping Unit (SKU). The classification of SKU is related to design evaluation, color demand forecast and size determination evaluation. In order to loose certain losses, which may be due to uncertain demand prediction, the Retail Planning Team has begun to implement the Management Information System (MIS) for the information search and forecast. The information search and forecast demand are usually done six months before the selling point. If an overview of the MIS is done it is seen that is it helpful in the assortment planning but a reliable systematic approach and a reliable conceptual model is rarely found in it. 1.5)

FASHION SELLING PERIOD

The buying as well as the selling cycle of merchandise depends upon the fashion cycle of consumer acceptance. However, predicting the right product cycle of an item is difficult, e.g.; the bestselling item of the last year may be the worse selling item of the current year if the product has a bad life cycle. Retail Planners in co-ordination with the Buyers depends on intuition with the prediction of demand for a fashion-sensitive product. The nature of fashion has a qualitative aspect in itself and hence the planning for an assortment has to be done keeping in mind the qualitative as well as the quantitative analysis methods, which would in turn result in accuracy in forecasting and reduce sales loss.

1.6)

RETAIL PLANNING PROCESS

The planning process is generally divided into three stages:  Merchandise Planning: To set up sales goals and inventory control system.  Assortment Planning: To decide quantity and quality of specific items.  Buying: Actual buying and rearrangement of the previous plans with vendors. (Done by the retail Buyers) 1.7) MERCHANDISE PLANNING  PLANNING SALES GOALS:


8

When buyers are purchasing merchandise it is necessary for them to control the purchase, hence retail planning comes into play which helps in providing direction and serves as a basis of control for any store. As a buyer, one must provide the right merchandise, at the right place, at the right time, in the right quantities, and at the right price. Hence to fulfill these goals the retail planning department helps to plan the merchandise budgets and merchandise assortments. The merchandise budget or the merchandise plan is a forecast of the specific merchandise purchase which covers a period of week, month, season and year which is known as: WTD, MTD, STD and YTD (Weekly, Monthly, Seasonal and Yearly Transaction Details) 1.8)

MERCHANDISE MANAGEMENT AND SALES FORECAST

The retail plan needs to be checked frequently in order to see whether the desired output is being achieved or not. Retailers need some type of planning and control device to guide their activity towards the achievement of their stated goals. Retail Planning is done with the help of a Micro strategy (MSTR), Management Information System (MIS). Sales forecast can be done in two ways:  Top-Down Planning: In this type of planning the top management decides the estimated sales for a given period and then distributing the sales to the individual department according to their past sales contribution. The top-down forecasting technique has a four step process:  Planning sales goals by reviewing past sales  Planning stock level for each order.  Planning the assortment plan by analyzing the sales potential for specific products.  Making a sales forecast report. 

Bottom-Up Planning:

The planned sales for each department are determined by department head and then the total sale is estimated by adding them.


9

This is the category MIS Report showing the sales. If we see this chart it is found that the sales of the Last Year(LY), Annual Budget Plan(ABP), and the Actual Sales(Act) for each category (mentioned in the first column) is taken and then the grand total sales is predicted by adding up the individual category, which as discussed earlier is the Bottom-Up Planning. 1.9)

MERCHANDISE PLAN

Key components of a merchandise plan include sales forecasts and stock planning. In addition, the amount of merchandise to be purchased each period to generate the planned sales is calculated. The six-month merchandise plan is the tool that translates profit objectives into a framework for merchandise planning and control. A merchandise plan in developed basically on two seasons:  

Autumn Winter Spring Summer

1.10) PURPOSE OF MERCHANDISE PLAN The merchandise plan regulates inventory levels in accordance with planned financial objectives. As with all merchandising activities, the essential goal of the merchandise plan is to minimize the use of capital and maximize profits. Key purposes of the merchandise plan are as follows:  

To provide an estimate of the amount of capital required to be invested in inventory for a specific period. To provide an estimate of planned sales for the period that translates into cash flow estimates for store management and accounting personnel. Graduation Project | BFT NIFT MUMBAI 2010-2014

10

1.11) PLANNING STOCK AND INVENTORY CONTROL The next step in planning process is to determine the amount of stock required to meet the customer demand. Volume of stock is calculated relative to desired sales and inventory turn. The amount of stock required is calculated with the help of Stock-To-Sales ratio. For the calculation of Stock-To-Sales ratio the GMROI% as well as the GM% is needed for the previous year. The following table will illustrate the following:  Calculation Of Stock-Sales Ratio: Store Name Month World

MC NSNT(*) Description

COGS(*) GMROI% GM% Stock Sales Ratio

PT-KOLKATA- April GARIAHAT ROAD

Women Ethnic

Salwar Kameez Set

445246.4 192195.7 131

56

2.33

PT-KOLKATA- April GARIAHAT ROAD

Women Ethnic

PN Fashion Basic Topwear

256625.9

51

2.15

124292.8 110

*NSNT: Net Sales Net Transactions *COGS: Cost of Goods Sold GMROI (Gross Margin Return on Investment) = {(NSNT-COGS)/COGS}*100% GM (Gross Margin) = {(NSNT-COGS)/NSNT}*100% STOCK TO SALES RATIO = GMROI%/GM% EXPLANATION Since the stock to sales ratio for the Salwar Kameez Set for PT-KOLKATA-GARIAHAT ROAD is 2.33 for April 2013 hence for April 2014 also the Stock to Sales ratio for that particular store and particular MC Description would be 2.33. Hence if the predicted sales for April 2014 is ‘x’ then the inventory available for that period would be x*2.33.


11

CHAPTER 2

LITERATURE REVIEW


12

2) LITERATURE REVIEW 2.1)

FORECASTING TECHNIQUE:

Forecasting techniques can be categorized in two broad categories: quantitative and qualitative. The techniques in the quantitative category include mathematical models such as moving average, time series method, straight line projection, exponential smoothing, regression, trend-line analysis, simulation, life-cycle analysis, decomposition, Box-Jenkins, expert systems, and neural network. The techniques in the qualitative category include subjective or intuitive models such as jury or executive opinion, sales force composite, and customer expectations Along with qualitative and quantitative, forecasting models can be categorized as time-series, causal, and judgmental. A time-series model uses past data as the basis for estimating future results. The models that fall into this category include decomposition, moving average, exponential smoothing, and Box-Jenkins. The premise of a causal model is that a particular outcome is directly influenced by some other predictable factor. These techniques include regression models. Judgmental techniques are often called subjective because they rely on intuition, opinions, and probability to derive the forecast. These techniques include expert opinion, Delphi, sales force composite, customer expectations (customer surveys), and simulation (Kress, G., 1985, Practical techniques of business forecasting, Westport) Typically, the two forms of forecasting error measures used to judge forecasting performance are mean absolute deviation (MAD) and mean absolute percentage error (MAPE). For both MAD and MAPE, a lower absolute value is preferred to a higher absolute value. MAD is the difference between the actual sales and the forecast sales, absolute values are calculated over a period of time, and the mean is derived from these absolute differences. MAPE is used with large amounts of data, and forecasters may prefer to measure error in percentage (Business Forecasting with Student CD [J. Holton Wilson, Barry Keating, Tata McGrawHill) Three planning horizons for forecasting exist. The short-term forecast usually covers a period of less than three months. The medium-term forecast usually covers a period of three months to two years. And, the long-term forecast usually covers a period of more than two years. Generally, the short-term forecast is used for the daily operation and plans of a company. The long-term forecast is used more for strategic planning. Forecasting systems for operations management: Stephen A. Delurgio and Carl D.Bhame, 1991, (Business One Irwin, Homewood, IL)


13

2.2)

FORECASTING AS A PART OF MANAGEMENT PROCESS

There can certainly be no more important activity in the business organization than the effective development of sales forecasts and application of these forecasts to the organization’s various functional needs. Closs, Oaks, & Wisdo (1989) argued that a sales forecast must incorporate 1. The correct use of forecasting techniques, 2. Forecasting systems that effectively interact with the corporate management information system, and 3. Recognition of the impact of forecasting management philosophy upon ultimate accuracy. A substantial gap still exists between applications and what is both desirable and obtainable. An examination of the forecasting and marketing literature suggests that a structure is needed for handling the issues that the practitioner must address. Forecasting Methods for Management by Spyros Makridakis and Steven C. Wheelwright (1977, Hardcover) Various functional areas or departments may need on-going information on forecasts and forecasting accuracy, even though they are not allowed to make changes to forecasts. The departments that are most often allowed to review forecasts are marketing, finance, production, sales, and planning. Having access to the sales forecast information as well as the ability to disseminate the information is important. Sales Forecasting Management: A Demand Management Approach By John T. Mentzer &

Mark A. Moon) Behavioral and organizational issues exist when integrating the forecasting system into a company. An important aspect of the behavior issue involves the interface between the preparer of forecasts and the users of forecasts. A need exists for a clear definition of tasks and priorities with regard to forecasting applications as well as a need for respect and understanding of each other's position. An important aspect of the organizational issue involves differences among the needs of each department that uses the forecast. Forecasting Methods for Management by Spyros Makridakis and Steven C. Wheelwright (1977, Hardcover)


14

Because the sales forecast is the bonding tool that draws together the different line and support functions, all of the components of the organization must use the same forecast and assumptions. A business organization is an integrated group of activities, which requires coordination and common goals to result in profit for the company. Evidence has shown that, if there is not a sufficient degree of acceptance of the forecast and its validity, the different functional areas will in fact develop their own independent forecasts. This has the obvious effect of creating chaos, inefficiency, and substantial additional costs. The conflict and chaos created by the use of different sales forecasts can be detrimental to the organization's efforts and have a variety of undesirable side effects, including high inventory levels, inventory obsolescence, over utilized or underutilized plants, and unnecessary facilities. These are serious consequences potentially costing the business millions of dollars in excess capitalization due to ineffective sales forecasting (“Forecasting in the 1990” by Lawless, 1990). 2.3)

PURPOSE OF FORECASTING

Several empirical studies focused on why businesses produce forecasts and the use they make of the latter. In White's (1986) survey, 64% of respondents regarded the purpose of a sales forecast as a goal setting device-a statement of desired performance; only 30% wanted to derive a true assessment of the market potential. This finding was independent of firm size. However, smaller firms used sales forecasts more often for personnel planning while for larger firms sales quota setting and purchasing planning were frequent uses. Mentzer and Cox (1984) enquired about the first, second and third most important areas of forecast usage. The majority of firms regarded production planning and budgeting as important decision areas, a finding also observed by Rothe (1978), McHugh and Sparkes (1983) and Peterson (1993). Peterson (1993) also observed among his sample of retailers that smaller firms used sales forecasts less frequently for planning purposes than larger firms, while Herbig (1994) found that industrial goods firms regarded the forecasting of industry trends, applications and technologies as being more important than did consumer goods firms. 2.4)

APPROACHES AND METHOD OF FORECASTING

Steen (1992) studied the importance of team-based forecasting. Both Steen (1992) and Kahn & Mentzer (1994) concluded that team-based forecasting tends to improve forecast accuracy, and managers are more satisfied when forecasts are developed with inter-functional efforts. According to Kahn & Mentzer (1994), four general approaches to sales forecasting exist.  The first approach is one in which each department develops and uses its own sales forecast. This is called the independent approach. Graduation Project | BFT NIFT MUMBAI 2010-2014

15

 The second approach has only one department responsible for developing the sales forecast. This is called the concentrated approach.  The third approach has a forecast team comprised of representatives from multiple departments responsible for developing the sales forecast. This is called the consensus approach.  Finally, the fourth approach has each department develop its own forecast, but a forecast team comprised of representatives from multiple departments is responsible for arriving at the final sales corporate forecast. This is called the negotiated approach. Approaches one and two are non-team-based approaches, while approaches three and four are team-based approaches. Sales Forecasting Management: A Demand Management Approach By John T. Mentzer &

Mark A. Moon) Gordon, Morris, & Dangerfield (1997) suggested two general approaches to forecasting: topdown (TD) and bottom-up (BU) approaches. In the top-down approach, data are used to develop a forecast, which is then desegregated into individual units based on their historical fraction of sales. The bottom-up approach allows each unit to prepare a separate forecast, which is aggregated. Gordon et al., (1997) concluded that the bottom-up approach outperforms the top-down approach in improving forecast accuracy. A goal programming model for hierarchical forecasting by Gordon, Morris, & Dangerfield (1997) Forecasts assist marketing managers improve decision-making. In an organizational design context, forecasting should not be regarded as a self-contained activity, but should be integrated within the planning context of which it is a part. Managerial evaluation of sales forecasting by Mahmoud, Rice, & Malhotra, 1988 When an organization has its own forecasting expertise (prepares its own forecasts) that expertise should not be separated into a self-contained department. Forecasting and planning functions should be combined. Involvement of the forecasters in planning enables them to select criteria for evaluating forecasting methods that are meaningful within the planning context. Forecasting methods for management by Steven C. Wheelwright, 1988


16

2.5)

ACCURACY IN FORECASTING

Accuracy of forecasts improves when the source of error is identified and corrected. The sales forecast is critically important in business because it is often the starting point for all operations or planning. Errors in forecasts have costs, which often are very high. These costs have direct effects on budgeting, planning, production, and perhaps prices. Despite the errors, forecasts must be conducted in order to make plans for the future. Many of the weaknesses of the sales forecasting system appeared to be related to organizational and structural problems. Due to lack of assigned areas of responsibility, the consensus forecasting approach worked to dilute forecasting responsibility. A lack of integration and agreement on control mechanisms across all sales forecasting activities and an absence of an agreed-upon mechanism to systematically gather sales force input into the forecasting process existed. Business forecasting methods by Jarrett, 1987, (Basil Blackwell, Oxford) 2.6)

INFORMATION SEARCH IN MERCHANDISE PLANNING

Clodfelter (1993) mentioned that various internal and external information sources are available to help forecasting consumer demand and selecting product line in an assortment plan.  The internal sources may be store records, merchandise plan report, and sales people’s opinions.  The external sources may include: (a) customer panel, (b) consumer magazines and trade publications, (c)vendor opinions, (d) trade associations, (e) competitors, (f) fashion forecasts magazines, and (g)reporting bureaus (i.e., demographic data) Kline and Wagner (1994) found that records of past sales had moderate effects on retail buyers’ decisions. Although the decision-making task involved new merchandise, with no selling history, selling records for established merchandise may have documented fashion trends and provided direction for buying new items. Retail Buying: From Staples to Fashion to Fads by Richard Clodfelter (Feb 1, 1993


17

CHAPTER 3

RESEARCH METHODOLOGY


18

3) RESEARCH METHODOLOGY 3.1)

MANUAL SALES FORECASTING:

Sales forecasting is an essential activity for a retail organization in order to plan the amount of merchandise which should be available in the stores for a particular time period in order to meet the customer demands.

Data Collection from past year’s sales of SS-13

Planning Sales Goals

Planning Inventory Control

Making a Merchandise Sales Calendar

Comparing the manual sales forecast with the MIS sales forecast

3.2) DATA COLLECTION: LAST YEAR’s SALES DATA FROM MIS: The last year’s sales data is obtained from a Micro-strategy (MSTR) MIS. MSTR is an MIS which is used by the Retail Planning department in order to get the sales of the merchandise. Through this an MC-wise sales report is extracted. MC-Wise Sales Planning Report: MC stands for Merchandise Category. Merchandise categories helps to classify and structure all aspects of the merchandise in the enterprise. In doing so, each article is assigned to a specific merchandise category. Now to understand MC let us consider two brands of pantaloons i.e. , ETHNICITY and ALL. Out of various products under these brands they have a product called as aLLLadiesMixnMatch. Now for both the brands for this product they are given a particular unique MC Code. Hence that particular MC Code would stand for that product only irrespective of whether it is made by some third brand. Graduation Project | BFT NIFT MUMBAI 2010-2014

19

The MIS looks as follows.

Here, entering previous year’s date for particular month would generate a report showing the last year sales as well as the last to last year’s sale. The data obtained here is in the raw form and hence it has to be mapped with a Master MC classification list in order to get the data for each category.

3.3)

PLANNING SALES GOALS

There are two broad categories of forecasting techniques: quantitative methods and qualitative methods. Quantitative methods are based on algorithms of varying complexity, while qualitative methods are based on educated guessing. I'll focus on quantitative methods here. QUANTITATIVE METHOD OF FORECASTING  Time Series Method Time series method has been used to make a forecast purely on historical patterns in the data. Like forecasting for the month of January 2014 will require the last year sales for the individual categories as well as the past few years data to come across the increment factor in the basic sales due to the increase in cost price which in turn results in the increment of the sales.


20

Time-series method is the most simplest and accurate, particularly over the short term. Most quantitative forecasting methods try to explain patterns in historical data as a means of using those patterns to forecast future patterns.

1) Getting the increment factor with respect to 2012 sales data

5) Calculating the planned Beginning Of Month inventory. (BOM)

2) Calculating the GM% as well as the GMROI%

3) Identifying the stock to sales ratio for a particular category

4) Arriving at the next year’s sales plan

SALES REPORT for 17-23rd FEB 2013 World Description Men Total Women Western Total Non Apps Total Women Ethnic Total Kids Total Grand Total

NSNT NSNT NSNT NSNT NSNT RGM (LY) GM (LY)% (2009) (2010) (2011) (2012) (2013) (2013) (2013) 346.25 370.54 390.65 422.12 468.13 193.98 46% 300.19 315.45 326.59 355.18 368.07 141.42 40% 231.47 219.64

249.98 235.74

268.36 248.23

285.32 261.42

318.81 293.94

80.46 116.22

28% 44%

84.25 89.35 1181.8 1261.06

94.26 1328.09

100.13 1,424.17

129.40 1,568.35

40.39 572.47

40% 40%

Grand total sales are as follows: YEAR 2009 2010 2011 2012 2013

SALES 1181.8 1261.06 1328.09 1424.17 1658.35

INCREMENT FACTOR 6.7 5.31 7.23 10.1


21

Calculating each category as a %age of total sales: 2009

2010

2011

2012


22

2013

It can be seen from the table that the increment factor for year 2010 is 6.7%, 2011 is 5.31%, 2012 is 7.23% and for year 2013 is 10.1%. Hence the mean increment would be, (6.7 + 5.31 + 7.23 + 10.1) %/4 = 7.33%, Therefore the grand total sales for week 17-23rd Feb 2014 would be: =1568.35 + 7.33% of 1568.35 = 1683.31 From the above pie charts it can be seen that the contribution of each category towards the total sales is almost same over the years hence taking the percentage contribution of each category for the last year to get the sales for each category as follows:

World Description

LY (NSNT)

Men Total Women Western Total Non Apps Total Women Ethnic Total Kids Total Grand Total

30% 23% 20% 19% 8%

ABP (Manual) 30% * 1683.31 23% * 1683.31 20% * 1683.31 19% * 1683.31 8% * 1683.31

504.993 387.1613 336.662 319.8289 134.6648 1683.31

Inventory required = Predicted Sales * Stock-to-sales ratio Stock-to-Sales ratio = GMROI/GM %( LY)


23

GMROI= [RGM(LY)/NSNT(LY)-RGM(LY)]*100 = (572.47/1568.35)*100 = 36.5% Therefore Stock-to-sales ratio = 36.5/29 = 1.25 Hence the planned stock for 17-23rd Feb 2014 = 1683.31* 1.25 = 2118.73 3.4)

BEGINNING OF WEEK INVENTORY (BOW):

The Annual Budget Plan for a particular month is the Beginning of Month inventory for that particular month. The BOM is the stock which is available for sale in that particular month.

3.5)

END OF WEEK INVENTORY (EOW):

The end of month inventory adds to the beginning of month inventory for the next month. The EOM is calculated by subtracting the net sales for that particular month from the beginning of month inventory for a particular period.

MEN’s Beginning Of Month Inventory ABP ACTUAL SALES (1723rd Feb 2014) End Of Month Inventory

Women Ethnic

30% * 2118.73 = 635.61 504.99 510.36

Women Non Apps Western 23% * 2118.73 = 20% * 487.3 2118.73 = 423.74 387.16 336.66 381.24 340.58

319.82 315.89

134.66 131.58

125.25

106.06

86.66

37.91

83.16

KIDS

19% * 2118.73 = 8% * 2118.73 402.55 =169.49

Similarly the forecasting is done for the next weeks. But one problem with this kind of forecasting is that it does not consider the product life cycle, launch of a new product or a new store. Suppose a new product is launched in the market, for that new product the past data is not available and hence forecast for that product is merely based on the intuition of the buyer. Due to which the forecast may not be correct. Hence some new forecasting methods will be applied which would deal with the trend analysis of the sales data for first few (5 – 6) weeks of a particular season and then modifying the forecast calendar which was made earlier. This method will be useful because if we observe the data for the first few weeks, it would cover all the aspects like current trend of a product, launch and acceptance of a new product, etc. Graduation Project | BFT NIFT MUMBAI 2010-2014

24

Since initially we do not have the actual sales value for SS-14 we would do our normal forecasting for SS-14. But after 5 or 6 weeks when we observe a trend in the sales we follow the following process for the sales forecasting for the upcoming weeks:


25

CHAPTER 4

IMPLEMENTATION AND DATA ANALYSIS Graduation Project | BFT NIFT MUMBAI 2010-2014

26

4.1)

PHASE I OF MIS AUTOMATION (BUILDING A SALES FORECASTING TECHNIQUE FOR MIS)

4.2)

FORECASTING TECHNIQUES

Forecasting methods can be classified as qualitative and quantitative. Qualitative methods generally involve the use of expert judgment to develop forecast. Such methods are applicable when the historical data on the variable being forecast are either not applicable or unavailable. In my project of MIS Automation, building a logic on the basis of judgment and opinions is not possible for the MIS team to project the future sales. Hence a proper mathematical logic has to be build which could be used in the form of coding in order to predict the sales. Hence a quantitative sales forecasting method has to be used. Quantitative sales forecasting methods can be used when: 1) The past information about the variable being forecast is available. 2) The information can be quantified. 3) It is reasonable to assume the pattern of the past will continue in the future. In such cases a forecast can be developed using a time series or a casual method.

Quantitative Approaches to Sales Forecasting:  Quantitative methods are based on an analysis of historical data concerning one or more time series.  A time series is a set of observations measured at successive points in time or over successive periods of time.  If the historical data used are restricted to past values of the series that we are trying to forecast, the procedure is called a time series method.  If the historical data used involve other time series that are believed to be related to the time series that we are trying to forecast, the procedure is called a causal method. Time-Series Method: If the historical data are restricted to past values of the variable to be forecast, the forecasting method is called as time-series method and the historical data are referred to as a time series. The objective of time series analysis is to discover a pattern in the historical data or time series and then extrapolate the pattern into the future; the forecast is based solely on past values of the variable and/or on past forecast errors. The pattern of the data is an important factor in understanding how the time series has behaved in the past. If such behavior can be expected to continue in the future, we can use the past pattern to guide us in selecting an appropriate forecasting method. Graduation Project | BFT NIFT MUMBAI 2010-2014

27

To identify the underlying pattern in the data, a useful first step is to construct a time series plot. A time series plot is a graphical presentation of the relationship between time and the time series variable; time is on the horizontal axis and the time series values are shown on the vertical axis. Let us review some of the common types of data patterns that can be identified when examining a time series plot. 4.3)

TIME SERIES-PLOT:

Let us consider the past 9 week’s sales for different categories of Pantaloons which are classified as MEN, WOMEN WESTERN, NON APPS, WOMEN ETHNIC and NON APPS. 1) CATEGORY: MENS WEEKS th

SEASON AW-13 AW-13 AW-13 SS-14 SS-14 SS-14 SS-14 SS-14 SS-14

nd

1 (27 Jan – 2 Feb) 2 (3rd Feb – 9th Feb) 3 (10th Feb – 16th Feb) 4 (17th Feb – 23rd Feb) 5 (24th Feb – 2nd March) 6 (3rd March – 9th March) 7 (10th March – 16th March) 8 (17th March – 23rd March) 9 (24th March – 30th March)

ACTUAL SALES ( In Lakh) 926.53 1012.59 1113.26 548.13 546.85 551.58 603.63 660.83 715.42

Time Series Plot (MEN) 1200 1000

SALES

800 600 400 200 0 0

2

4

6

8

10

WEEKS


28

2) CATEGORY: WOMEN WESTERN WEEKS th


nd



Time Series Plot (WOMEN WESTERN) 700 600

SALES

500 400 300 200 100 0 0

2

4

6

8

10

WEEKS 3) CATEGORY: NON APPS WEEKS th

nd





29

Time Series Plot (NON APPS) 700 600

SALES

500 400 300 200 100 0 0

2

4

6

8

10

WEEKS 4) CATEGORY: WOMEN ETHNIC WEEKS th


nd



Time-Series Plot (WOMEN ETHNIC) 600 500

SALES

400 300 200 100 0 0

2

4

6

8

10

WEEKS


30

5) CATEGORY: KIDS WEEKS th


nd


ACTUAL SALES ( In Lakh) 203.43 218.96 255.21 129.4 128.4 147.78 200 218.41 250.61

Time-Series Plot (KIDS) 300 250

SALES

200 150 100 50 0 0

2

4

6

8

10

WEEKS

4.4)

ANALYSIS OF THE ABOVE TIME SERIES PLOT:

It is observed that from all the above time series curve only the Women Ethnic category shows a different pattern which is known as horizontal pattern of trend analysis in which the sales values keeps on moving around an average value of all the sales figure. The pattern of sales keeps on increasing and decreasing as the weeks keep on moving. Hence for these type of sales trends the forecasting technique that are used are described below. For rest of the categories there could be seen a particular trend of constant rising in the sales values. Hence for those categories a forecasting tool called as Linear Regression is used in order for sales forecasting.


31

4.5)

SELECTING A FORECASTING METHOD:

The underlying pattern in the time series is an important factor in selecting a forecasting method. Thus, a time series plot should be one of the first things developed when trying to determine what forecasting method to use. If we see a horizontal pattern, then we need to select a method appropriate for this type of pattern. Similarly, if we observe a trend in the data, then we need to use a forecasting method that has the capability to handle trend effectively. Now as I have discussed earlier that the Women Ethnic category shows a horizontal pattern in the time series plot hence a forecasting technique appropriate for horizontal data pattern will be used. 4.6)

FORECAST ACCURACY:

In this section let me begin by developing forecast for the women ethnic category using the simplest of all forecasting methods: an approach that uses the most recent week’s sales value as a forecast for next week. For example, the actual sales for week 4 (293.94 Lakh) is used as the sales forecast for the week 5 and the actual sales for the week 5 (320.27) is used as the sales forecast for the week 6 and so on. Because of the simplicity of this method it is known as the naïve forecasting method. The main question which arises is that how accurate are the forecasts obtained using this naive forecasting method? For answering this question several measures of forecast accuracy are checked upon. These measures are used to determine how well a particular forecasting method is able to reproduce the time series data that are already available. By selecting the method that has the best accuracy for the data already known, we hope to increase the likelihood that we will obtain better forecasts for future time periods. The key concept associated with measuring forecast accuracy is forecast error, defined as:

Forecast Error = Actual Sales - Forecast

4.7) NAÏVE METHOD OF SALES FORECASTING: In this type of sales forecasting method the actual sales for the previous period is taken as the forecasted sales for the next period. In other words the most recent sales values are taken as the forecast for upcoming period.


32

COMPUTING FORECASTS AND MEASURES OF FORECAST ACCURACY USING THE MOST RECENT VALUE AS THE FORECAST FOR THE NEXT PERIOD Week

4 (17th Feb – 23rd Feb) 5 (24th Feb – 2nd March) 6 (3rd March – 9th March) 7 (10th March – 16th March) 8 (17th March – 23rd March) 9 (24th March – 30th March)

TOTALS

Actual Sales

Forecast

Forecast Error

Absolute Value Of Forecast Error

Squared Forecast Error

Percentage Absolute Error Value of Percentage Error

26.33

26.33

693.26

8.22

8.22

73.4

73.4

5387.56

18.64

18.64

-93.59

93.59

8759.08

-31.18

31.18

59.87

59.87

3584.41

16.63

16.63

-12.76

12.76

162.81

-3.6

3.6

53.25

265.95

18587.12

8.71

78.27

293.94 320.27

393.67

300.08

359.95

347.19

293.94

320.27

393.67

300.08

359.95

The fact that the forecast error is positive indicates that in week 5 the forecasting method Under estimated the actual value of sales, whereas in week 7 as well as week 9 the forecast error is negative which indicates that for both these weeks the forecast made is higher than the actual sales. A simple measure of forecast accuracy is the mean or average of the forecast errors. The Table above shows that the sum of the forecast errors as 53.25 thus, the mean or average forecast error is: 53.25/5 = 10.65 Note that although the Women Western time series consists of 6 values, to compute the mean error we divided the sum of the forecast errors by 5 because there are only 5 forecast errors. Because the mean forecast error is positive, the method is under forecasting; in other words, the observed values tend to be greater than the forecasted values. Because positive and


33

negative forecast errors tend to offset one another, the mean error is likely to be small thus, the mean error is not a very useful measure of forecast accuracy. The mean absolute error, denoted MAE, is a measure of forecast accuracy that avoids the problem of positive and negative forecast errors offsetting one another. MAE is the average of the absolute values of the forecast errors. The table shows that the sum of the absolute values of the forecast errors is 265.95 thus, MAE = average of the absolute value of forecast errors = 265.95/5 = 53.19 Another measure that avoids the problem of positive and negative forecast errors offsetting each other is obtained by computing the average of the squared forecast errors. This measure of forecast accuracy, referred to as the mean squared error, is denoted MSE. From the table the sum of squared errors is 18587.12 MSE = average of the sum of squared forecast errors = 18587.12/5 = 3717.42

The size of MAE and MSE depends upon the scale of the data. As a result, it is difficult to make comparisons for different time intervals, such as comparing a method of forecasting monthly sales to a method of forecasting weekly sales, or to make comparisons across different time series. To make comparisons like these we need to work with relative or percentage error measures. The mean absolute percentage error, denoted MAPE, is such a measure. To compute MAPE we must first compute the percentage error for each forecast. For example, the percentage error corresponding to the forecast of 293.94 in week 5 is computed by dividing the forecast error in week 5 by the actual value in week 5 and multiplying the result by 100. For week 5 the percentage error is computed as follows: Percentage error for week 5 = (293.94/320.27) * 100 = 8.22% Thus, the forecast error for week 5 is 8.22% of the observed value in week 5. A complete summary of the percentage errors is shown in the table in the column labeled Percentage Error. In the next column, we show the absolute value of the percentage error. The table shows that the sum of the absolute values of the percentage errors is 78.27 thus, MAPE = average of the absolute value of percentage forecast errors = 78.27/5 = 15.65% Summarizing the above result of forecasting: Naïve Method of Sales Forecasting (Women Ethnic) Forecasting Errors Value MAE 53.19 MSE 3717.42 MAPE 15.65 Graduation Project | BFT NIFT MUMBAI 2010-2014

34

4.8)

AVERAGE SALES METHOD:

COMPUTING FORECASTS AND MEASURES OF FORECAST ACCURACY USING THE AVERAGE OF ALL THE HISTORICAL DATA AS THE FORECAST FOR THE NEXT PERIOD

Week 4 (17th Feb – 23rd Feb) 5 (24th Feb – 2nd March) 6 (3rd March – 9th March) 7 (10th March – 16th March) 8 (17th March – 23rd March) 9 (24th March – 30th March)

Actual Sales

Forecast

Forecast Error



Absolute Percentage Value of Error Percentage Error

293.94

320.27

293.94 26.33

26.33

693.2689

8.221188

8.221188

307.105

86.565

86.565

7493.499

21.98923

21.98923

335.96

-35.88

35.88

1287.374

-11.9568

11.95681

326.99

32.96

32.96

1086.362

9.156827

9.156827

333.582

13.608 123.58

13.608 195.34

185.1777 10745.68

3.919468 31.32

3.919468 55.24

393.67

300.08

359.95

347.19

TOTALS

This method of sales forecasting is known as the Average of past sales. Suppose we use the average of all the historical data available as the forecast for the next period. We begin by developing a forecast for week 5. Since there is only one historical value available prior to week 5, the forecast for week 5 is just the time series value in week 1, thus the forecast for week 5 is 293.34 lakhs. To compute the forecast for week 6, we take the average of the sales values in weeks 4 and 5. Thus, Forecast for week 6 =

293.94+320.27 2

Similarly, the forecast for week 7 is: Graduation Project | BFT NIFT MUMBAI 2010-2014

35

Forecast for week 7 =

293.94+320.27+393.67 3

The forecasts obtained using this method for the women western category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

MSE =

195.34 5

10745 .68

MAPE =

5 55.24 5

= 39.06 = 2149.13 = 11.04

We can now compare the accuracy of the two forecasting methods we have considered in this section by comparing the values of MAE, MSE, and MAPE for each method.

MAE MSE MAPE

Naïve Method 53.19 3717.42 15.65

Average Sales Method 39.06 2149.13 11.04

4.9) MOVING AVERAGES AND EXPONENTIAL SMOOTHNING In this section I will discuss three forecasting methods that are appropriate for a time series with a horizontal pattern:  Moving averages  Weighted moving averages and  Exponential smoothing. These methods also adapt well to changes in the level of a horizontal pattern. However, without modification they are not appropriate when significant trend, cyclical, or seasonal effects are present. Because the objective of each of these methods is to “smooth out” the random fluctuations in the time series, they are referred to as smoothing methods. These methods are easy to use and generally provide a high level of accuracy for short-range forecasts, such as a forecast for the next time period.


36

4.10) MOVING AVERAGES: The moving averages method uses the average of the most recent k data values in the time series as the forecast for the next period. Mathematically, a moving average forecast of order k is as follows:

MOVING AVERAGE FORECAST OF ORDER k

Ft+1 =

𝑆𝑈𝑀 (𝑚𝑜𝑠𝑡 𝑟𝑒𝑐𝑒𝑛𝑡 𝑘 𝑑𝑎𝑡𝑎 𝑣𝑎𝑙𝑢𝑒𝑠 ) 𝑌𝑡+𝑌𝑡−1+⋯.+𝑌𝑡−𝑘+1 =

𝑘

𝑘

where,

Ft+1 = forecast of time series for period t+1 Yt = actual value of the time series in period t

THREE WEEK MOVING AVERAGE:

Week 4 (17th Feb – 23rd Feb) 5 (24th Feb – 2nd March) 6 (3rd March – 9th March) 7 (10th March – 16th March) 8 (17th

Actual Sales

Forecast Error



26.33

26.33

693.2689

8.221188

8.221188

307.105

86.565

86.565

7493.499

21.98923

21.98923

335.96 338.0067

-35.88 21.94333

35.88 21.94333

1287.374 481.5099

-11.9568 6.096217

11.95681 6.096217

Forecast


293.94

320.27

293.94

393.67

300.08 359.95


37

March – 23rd March) 9 (24th March – 30th March)

TOTALS

347.19 351.2333

-4.04333 94.91

4.043333 174.76

16.34854 9972

-1.16459 23.18

1.164588 49.42

The term moving is used because every time a new observation becomes available for the time series, it replaces the oldest observation in the equation and a new average is computed. As a result, the average will change, or move, as new observations become available. To illustrate the moving averages method, let us return to the sales of women western category which has a horizontal pattern in time series. Thus, the smoothing methods of this section are applicable. To use moving averages to forecast a time series, we must first select the order, or number of time series values, to be included in the moving average. If only the most recent values of the time series are considered relevant, a small value of k is preferred. If more past values are considered relevant, then a larger value of k is better. As mentioned earlier, a time series with a horizontal pattern can shift to a new level over time. A moving average will adapt to the new level of the series and resume providing good forecasts in k periods. Thus, a smaller value of k will track shifts in a time series more quickly. But larger values of k will be more effective in smoothing out the random fluctuations over time. So managerial judgment based on an understanding of the behavior of a time series is helpful in choosing a good value for k. To illustrate how moving averages can be used to forecast sales, we will use a three-week moving average (k = 3). We begin by computing the forecast of sales in week 5 which is the actual sales in week 4. For week 6 the forecast is done by taking the average of week 4 and 5 sales. For week 7 onwards:

F7 = average of weeks 4-6 =

F8 = average of weeks 5-7 =

293.94+320.27+393.67 3

= 335.96

320.27+393.67+300.08 3

= 338

The forecasts obtained using this method for the women western category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained:


38

MAE =

MSE =

174.76 5 9972

MAPE = Naïve Method MAE MSE MAPE

53.19 3717.42 15.65

5

= 1994.4

49.42 5

= 34.95

= 9.88 Average Sales Method 39.06 2149.13 11.04

Moving Average Method 34.95 1994.4 9.88

4.11) EXPONENTIAL SMOOTHNING: Exponential smoothing also uses a weighted average of past time series values as a forecast, it is a special case of the weighted moving averages method in which we select only one weight— the weight for the most recent observation. The weights for the other data values are computed automatically and become smaller as the observations move farther into the past. The exponential smoothing equation follows.

Ft+1 = aYt + (1-a)Ft where,

Ft+1 = forecast of time series for period t+1 Yt = actual value of time series in period t Ft = forecast of time series for period t a = smoothing constant (0 <=a <= 1)

The above relation shows that the forecast for period t + 1 is a weighted average of the actual value in period t and the forecast for period t. The weight given to the actual value in period t is the smoothing constant α and the weight given to the forecast in period t is 1 – α. Graduation Project | BFT NIFT MUMBAI 2010-2014

39

It turns out that the exponential smoothing forecast for any period is actually a weighted average of all the previous actual values of the time series. Let us illustrate by working with a time series involving only three periods of data: Y1, Y2, and Y3. To initiate the calculations, we let F1 equal the actual value of the time series in period 1, that is, F1 = Y1. Hence, the forecast for period 2 is

F2 = aY1 + (1-a)F1 = aY1 + (1-a)Y1 = Y1 We see that the exponential smoothing forecast for period 2 is equal to the actual value of the time series in period 1. The forecast for period 3 is:

F3 = aY2 + (1-a)F2 = aY2 + (1-a)Y1 Finally, substituting this expression for F3 in the expression for F4, we obtain

F4 = aY3 + (1-a)F3 = aY3 + (1-a)[ aY2 + (1-a)Y1] = aY3 + a(1-a)Y2 + (1-a)2Y1 We now see that F4 is a weighted average of the first three time series values. The sum of the coefficients, or weights, for Y1, Y2, and Y3 equals 1. A similar argument can be made to show that, in general, any forecast Ft+1 is a weighted average of all the previous time series values. Despite the fact that exponential smoothing provides a forecast that is a weighted average of all past observations, all past data do not need to be saved to compute the forecast for the next period. In fact, equation shows that once the value for the smoothing constant α is selected, only two pieces of information are needed to compute the forecast Yt, the actual value of the time series in period t, and Ft, the forecast for period t.  To illustrate the exponential smoothing approach, let us consider the women ethnic sales. As indicated previously, to start the calculations we set the exponential smoothing forecast for period 2 equal to the actual value of the time series in period 1. Thus, with Y1 = 293.94, we set F2 = 293.94 to initiate the computations. Referring to the time series data we find an actual time series value in period 2 of Y2 = 320.94 Continuing with the exponential smoothing computations using a smoothing constant of α = .2, we obtain the following forecast for period 3:

F3 = .2Y2 + .8F2 = .2(320.27) + .8(293.94) = 320.27 Graduation Project | BFT NIFT MUMBAI 2010-2014

40

SUMMARY OF THE EXPONENTIAL SMOOTHING FORECASTS AND FORECAST ERRORS FOR THE WOMEN ETHNIC TIME SERIES WITH SMOOTHING CONSTANT α = .2

Week 4 (17th Feb – 23rd Feb) 5 (24th Feb – 2nd March) 6 (3rd March – 9th March) 7 (10th March – 16th March) 8 (17th March – 23rd March) 9 (24th March – 30th March)

TOTALS

Actual Sales

Forecast


Forecast Error



293.94

320.27

293.94 26.33

26.33

693.2689

8.221188

8.221188

320.27

73.4

73.4

5387.56

18.64506

18.64506

318.09

-18.01

18.01

324.3601

-6.00173

6.001733

314.48

45.47

45.47

2067.521

12.63231

12.63231

323.54

23.65 150.84

23.65 186.86

559.3225 9032

6.811832 40.3

6.811832 52.31

393.67

300.08

359.95

347.19

The forecasts obtained using this method for the women western category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

186.86 5 9032

= 37.37

= 1806 5 Graduation Project | BFT NIFT MUMBAI 2010-2014 41

MSE =

MAPE =

52.31 5

= 10.46

4.12) COMPARISON OF VARIOUS FORECASTING TECHNIQUES TO FIND OUT THE BEST LOGIC FOR THE MIS Naïve Method MAE MSE MAPE

53.19 3717.42 15.65

Average Sales Method 39.06 2149.13 11.04

Moving Average Method 34.95 1994.4 9.88

EXPONENTIAL SMOOTHING 37.37 1806 10.46

From the above chart it can be seen that the Moving Average Method is showing the least MAPE as well as the MAE and the MSE is also very low as compared to the Naïve as well as the Average sales method. Hence for the MIS Automation of sales forecasting for the women ethnic category the sales forecasting tool which will be used is the Moving Average Method.

4.13) TREND PATTERN OF SALES FORECASTING: The trend pattern of sales forecasting is used where we can see a general trend like continuous rising or continuous decrease in the sales of the product categories. Earlier through the time series plot it has been shown that all the categories except the Women Western category shows a trend pattern in the sales data and hence for the forecasting of sales for those categories trend analysis is used. The method used for trend pattern forecasting is linear trend regression. 4.14) LINEAR TREND REGRESSION FOR MEN’s CATEGORY: 1) Let’s first of all start with the Men’s Category. LINEAR TREND EQUATION

Tt = b0 + b1t Where,

Tt = linear trend forecast in period t b0 = intercept of the linear trend line b1 = slope of the line trend line t = time period


42

WEEKS th

SEASON SS-14 SS-14 SS-14 SS-14 SS-14 SS-14

rd

1 (17 Feb – 23 Feb) 2 (24th Feb – 2nd March) 3 (3rd March – 9th March) 4 (10th March – 16th March) 5 (17th March – 23rd March) 6 (24th March – 30th March)

ACTUAL SALES ( In Lakh) 548.13 546.85 551.58 603.63 660.83 715.42

In the above table the time variable begins at t=1 corresponding to the first time series observation and continues until t=6 corresponding to the most recent time series observation. Formulas for computing the excessive regression coefficients (b 0 and b1) are:

COMPUTING THE SLOPE AND INTERCEPTFOR A LINEAR TREND 𝑠𝑢𝑚𝑜𝑓 𝑡=1𝑡𝑜 6 𝑡−𝑡 ^ [Yt−𝑌^]

B1=

𝑠𝑢𝑚 𝑜𝑓 𝑡=1𝑡𝑜 6 𝑡−𝑡 ^ 2

B0 = Y^-b1t^ Where, Yt = value of the time series in period t Y^ = average value of the time series t^ = average value t

To compute the linear trend equation for the men’s category time series, we begin the calculations by computing and using the information in Table above: t^ = 21/6 = 3.5 Y^ = 3626.44/6 = 604.4 Using these values we can compute the slope and intercept of the trend line:


43

t 1 2 3 4 5 6 21

TOTALS

Yt 548.13 546.85 551.58 603.63 660.83 715.42 3626.44

t-t^ -2.5 -1.5 -0.5 0.5 1.5 2.5

Yt-Y^ -56.27 -57.55 -52.82 -0.77 56.43 111.02

(t-t^)(Yt-Y^) 140.675 86.325 26.41 -0.385 84.645 277.55 615.22

(t-t^)2 6.25 2.25 0.25 0.25 2.25 6.25 17.5

From the above table, B1 = 615.22/17.5 = 35.15 B0 = 604.4 – 35.15(3.5) = 487.37

SUMMARY OF THE LINEAR TREND FORECASTS AND FORECAST ERRORS FOR THE MEN’s TIME SERIES

Actual Sales

Week 1 2 3 4 5 6

548.13 546.85 551.58 603.63 660.83 715.42

Forecast 522.52 557.67 592.82 627.97 663.12 698.27

Absolute Absolute Squared Value Of Percentage Value of Forecast Forecast Error Percentage Error Error Error 25.61 655.8721 4.672249 4.672249 10.82 117.0724 1.978605 1.978605 41.24 1700.738 7.476703 7.476703 24.34 592.4356 4.032271 4.032271 2.29 5.2441 0.346534 0.346534 17.15 294.1225 2.397193 2.397193 121.45 3365.48 20.9 20.9

Forecast Error 25.61 -10.82 -41.24 -24.34 -2.29 17.15

TOTALS

The forecasts obtained using this method for the Men’s category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

MSE =

121.45 6

3365 .48

MAPE =

6 20.9 6

= 20.24 = 560.91 = 3.48


44

Linear Regression Method of Sales Forecasting (Men’s) Forecasting Errors Value MAE 20.4 MSE 560.91 MAPE 3.48

4.15) LINEAR REGRESSION FOR SALES FORECASTING FOR NON APPS:

WEEKS th


rd



To compute the linear trend equation for the men’s category time series, we begin the calculations by computing and using the information in Table above: t^ = 21/6 = 3.5 Y^ = 2417/6 = 402.83 Using these values we can compute the slope and intercept of the trend line: t

TOTALS

1 2 3 4 5 6 21

Yt 318.81 330.81 331.53 425.9 492.77 517.18 2417

t-t^ -2.5 -1.5 -0.5 0.5 1.5 2.5

Yt-Y^ -84.02 -72.02 -71.3 23.07 89.94 114.35

(t-t^)(Yt-Y^) 210.05 108.03 35.65 11.535 134.91 285.875 786.05

(t-t^)2 6.25 2.25 0.25 0.25 2.25 6.25 17.5

From the above table, B1 = 786.05/17.5 = 44.91 B0 = 402.83 – 44.91(3.5) = 263.6 Graduation Project | BFT NIFT MUMBAI 2010-2014

45

SUMMARY OF THE LINEAR TREND FORECASTS AND FORECAST ERRORS FOR THE NON APPS SALES TIME SERIES

Week 1 2 3 4 5 6 TOTALS

Actual Sales 318.81 330.81 331.53 425.9 492.77 517.18

Forecast 308.55 353.46 398.37 443.28 488.19 533.1


Forecast Error 10.26 -22.65 -66.84 -17.38 4.58 -15.92

The forecasts obtained using this method for the Non Apps category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

MSE =

137.63 6

5662 .36

MAPE =

6 38.31 6

= 22.93 = 943.72 = 6.38

Linear Regression Method of Sales Forecasting (Non Apps) Forecasting Errors Value MAE 22.93 MSE 943.72 MAPE 6.38


46

4.16) LINEAR REGRESSION FOR SALES FORECASTING FOR WOMEN WESTERN:

WEEKS th


rd



To compute the linear trend equation for the men’s category time series, we begin the calculations by computing and using the information in Table above: t^ = 21/6 = 3.5 Y^ = 2888.49/6 = 481.41 Using these values we can compute the slope and intercept of the trend line: t

TOTALS

1 2 3 4 5 6 21

Yt 368.07 378.94 445.81 491.07 587.36 617.24 2888.49

t-t^ -2.5 -1.5 -0.5 0.5 1.5 2.5

Yt-Y^ -113.34 -102.47 -35.6 9.66 105.95 135.83

(t-t^)(Yt-Y^) 283.35 153.705 17.8 4.83 158.925 339.575 958.15

(t-t^)2 6.25 2.25 0.25 0.25 2.25 6.25 17.5



47

SUMMARY OF THE LINEAR TREND FORECASTS AND FORECAST ERRORS FOR THE WOMEN WESTERN SALES TIME SERIES


Actual Sales 368.07 378.94 445.81 491.07 587.36 617.24

Forecast 344.53 399.28 454.03 508.78 563.53 618.28


Forecast Error 23.54 -20.34 -8.22 -17.71 23.83 -1.04

The forecasts obtained using this method for the Women Western category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

MSE =

94.68 6

= 15.78

1918.01

MAPE =

6 21.43 6

= 319.66 = 3.57

Linear Regression Method of Sales Forecasting (WOMEN WESTERN) Forecasting Errors Value MAE 15.78 MSE 319.66 MAPE 3.57


48

4.17) LINEAR REGRESSION FOR SALES FORECASTING FOR KIDS CATEGORY:

WEEKS th


rd


ACTUAL SALES ( In Lakh) 129.4 128.4 147.78 200 218.41 250.61

To compute the linear trend equation for the men’s category time series, we begin the calculations by computing and using the information in Table above: t^ = 21/6 = 3.5 Y^ = 1074.6/6 = 179.1

Using these values we can compute the slope and intercept of the trend line: t

TOTALS

1 2 3 4 5 6 21

Yt 129.4 128.4 147.78 200 218.41 250.61 179.1

t-t^ -2.5 -1.5 -0.5 0.5 1.5 2.5

Yt-Y^ -49.7 -50.7 -31.32 20.9 39.31 71.51

(t-t^)(Yt-Y^) 124.25 76.05 15.66 10.45 58.965 178.775 464.15

(t-t^)2 6.25 2.25 0.25 0.25 2.25 6.25 17.5



49

SUMMARY OF THE LINEAR TREND FORECASTS AND FORECAST ERRORS FOR THE KIDS SALES TIME SERIES


Actual Sales 129.4 128.4 147.78 200 218.41 250.61

Forecast 112.8 139.32 165.84 192.36 218.88 245.4

Forecast Error 16.6 -10.92 -18.06 7.64 -0.47 5.21


The forecasts obtained using this method for the Women Western category are shown in the above table in the column labeled Forecast. Using the results shown in table, the following values of MAE, MSE, and MAPE are obtained: MAE =

MSE =

59.9 6

806.7

MAPE =

6

= 9.98 = 134.45

39.66 6

= 6.61

Linear Regression Method of Sales Forecasting (KIDS) Forecasting Errors Value MAE 9.98 MSE 134.45 MAPE 6.61


50

4.18) PHASE II OF THE MIS AUTOMATION: The phase II of the MIS automation deals with the automized sales report generation in order to get the sales performance for a particular merchandise for that particular week with respect to certain factors like the last year sales for that particular week or the difference between the sales as well as the budget planned. MIS is a Management Information System which deals in generating an automated sales report consisting of the WTD, MTD, STD as well as the YTD (Week, Month, Season and Year Till Date Transactions). The output is generated every week on Monday on the basis of some input data which are as follows:  Article Wise Sales Data for a particular week: Every product is assigned to a certain article ID. The article wise sales report is the base data which is required to generate the MIS Sales report. This article wise sales report is extracted from a MSTR (Micro Strategy) ERP system which contains the sales data associated with that particular article.

 ABP(Annual Budget Plan) and RGM(Rupee Gross Margin) date wise for a particular month: The ABP is the Annual Budget Plan which is basically the sales forecasting that has been discussed in the PHASE I of the MIS Automation. Initially the ABP was provided by the planning team but as already discussed the ABP has also been automized. The RGM is the Rupee Gross Margin which is the difference between the cost of goods sold which is the basic cost and the actual sales value. Graduation Project | BFT NIFT MUMBAI 2010-2014

51

 Article Hierarchy: Each product that is assigned to a particular article ID contains certain MC code against itself. Article Hierarchy is a master file which contains the MC-Codes along with the World, Type, Division, Brand and MC Description. The article wise sales which is extracted from the MSTR is only on the basis of article ID and does not contain the detailed description of the product hence to categorize the sales in the category of Men’s, Women Western, Non Apps, Women Ethnic and Kids the article hierarchy is needed.

 Space Master: The space master consists of the total area of the pantaloons store - wise in square foot. It is required in order to calculate the SSPD which is Sales per Square Foot Per Day. For example, one category of pantaloons is Women Ethnic under which there is a brand named as Akkriti have a salable merchandise as Ethnic young. So for that particular brand which is the space allocated for the different pantaloons stores is the space master.


52

4.19) DESIGNING THE LAYOUT OF THE MIS SALES REPORT:


53

 The cells marked in green are the columns which has to be included in the MIS Reports whereas the cells which are white are the logics on the basis of which the columns has to be developed.

4.20) BUILDING THE LOGICS FOR THE CALCULATED FIELDS IN THE MIS REPORT:  NSNT: Net Sales Value – Total Tax Amount: NSNT is the Net Sales Nil Tax. It is the sales value which does not contain the tax amount in it. From the article wise sales we get two separate columns, one is the Net Sales Value which is the total value of a merchandise including the Tax and another column which is known as the Total Tax amount. Hence subtracting Total Tax Amount from the Net Sales Value will give the NSNT.  Gr% over LY = (Act Sales-LY Sales)/LY sales *100 %: Growth %age over the last year denotes that by how much the merchandize sales value or volume has increased over the last year. A positive sign indicates that the business for a particular product has made a growth over the last year whereas a negative sign indicates that the business has been in a loss.  ASP = Sales Value / Sales Quantity: Suppose for the kids category we have to find out the average selling price, then for that we have to take the total sales figure of the kids category as well as the total quantity of all the merchandize available in the kids category. Hence the average selling price for a merchandize in the kids category is the total sales value divided by the total sales quantity.


54

 RGM Act = NSNT – Cost of Goods Sold: RGM which is basically the difference between the NSNT and the COGS is the gross profit the product has made over the cost of the production.  RGM% = RGM Act/NSNT*100  Intake Margin% ACT = Gross Sales Value(GSV)-Cost of Goods Sold/GSV*100 The initial markup which is applied to a product over the cost of the goods before applying the Markdown. The price of the product at this stage is called as the Gross Sales Value. The GSV – COGS will actually give the Intake margin amount. Markdowns are applied to this GSV only and after the markdown is applied it is known as the Net Sales Value. Hence if a product is sold at its GSV without the markdown being applied then the GSV is same the Net Sales Value.  Markdown% = Markdown Amount/GSV*100 Generally in order to attract the customers the merchandiser applies certain markdown on the product which is applied on the intake margin. Markdown %ages are always applied by taking into consideration the profit margin of the produst.  SSPD(Sales Per Square foot per day): NSNT/(7*area)

4.21) COLOR CODE LOGIC 

ALL Growth over Last Year

Great than 10% = Green 0-10% = Yellow Less than 0% = RED 

ALL LTL

Great than 5% = Green 0-5% = Yellow Less than 0% = RED 

ALL ABP

Great than equal 100% = Green 90-100% = Yellow Less than 90% = RED  GM LY% & GM BUDGET % - NO color Code  ALL GM ACT%

If actual Great than budget = Green Graduation Project | BFT NIFT MUMBAI 2010-2014

55

If actual is 95% of budget = Yellow If actual is less than 95% of budget = RED  

IM LY% - No Color IM ACT%

If actual Great than IM LY = Green If actual is 95% of IM LY = Yellow If actual is less than 95% of IM LY = RED  

MD LY% - No Color MD ACT%

If actual Less than MD LY = Green If actual is 100 to 105% of MD LY = Yellow If actual is greater than 105% of MD LY = RED

4.22) MIS SALES REPORT:

 Once all the data is gathered it is given to the MIS team which generates the sales report for that particular week.  Since every week there are some new MC codes which are created hence those MC codes do not have a classification of the World, Type, Division, Brand, and Product description hence for that particular week the sales that for those MC Codes are classified as not defined and for future prospects those MC’s needs to be classified so that the sales figure for each world description is defined.  After the MIS report is generated it has to be validated with the base data that is provided to the MIS team  The base data which is the Article wise sales report has to be looked up with MIS Master in order to arrange it in the form of World, Type and Division.  There are certain fields like the ABP, RGM the current week sales figure as well as the quantity which has to matched directly with the base data, where as there are some calculated fields like the Growth over last year, the Average Selling Price(ASP), GM%, IM%, and the MD%.


56

4.23) THE MIS REPORT OF PHASE II LOOKS AS FOLLOWS:


57

4.24) PHASE III OF THE MIS AUTOMATION: The phase III is an addition in the existing MIS Automated Report which would display a few new parameters like the Bought Options, Bought Quantity, Sell-through target, etc. The base data required for the Phase II are:  Season Master: A season master is a master file which contains the season name along with the season description. A stock report which shows the current store as well as warehouse stock available contains this season stock as well as old season stock. Hence this master file is needed to classify the season stocks separately.

 Site Master A site master contains a detailed description about the Pantaloons sites that whether it is a store or a warehouse, whether the site is currently active. A site master is used to classify the store as well as the warehouse stock quantity.


58

 PO-GRN Report: The PO-GRN report is used to find out the bought quantity of the goods that are made in a particular time period.  Stock Report: The stock report takes into consideration that how much stock is available in the stores as well as the warehouse. The stock report contains both this year season store stock quantity as well as last year season store stock quantity.  Prepack Article Color and Quantity: Out of all the given parameters the first four is a one-time data whereas the PO-GRN, Stock and prepack article color and quantity has to be given every week. PO-GRN and Stock report has to be extracted from the MSTR whereas the prepack color and quantity is derived from the SAP. For the Prepack color and quantity first of all the prepack article has to found out from the stock report and then a Prepack-BOM relation has to be derived. From the BOM a BOMComponent-Quantity relation has to be taken out. Finally for a particular component its color has to be derived.


59

STEPS TO GET THE ABOVE TABLE: Prepack-BOM: Table Code:

se16

Table Name: mast


60

Enter Table Code se16


61

Enter Table Name: mast

CLICK HERE

Clear these


62

1) Then Click here

2) Click here to pull the prepack codes saved in notepad

Click here to execute


63

Click here to export in excel


64

 Save the BOM generated in Notepad file. 2) BOM-Component-Quantity Table Code:

se16

Table Name:

stop

Click here and the pull the BOM text file. Rest of the procedure is same.


65

 Make a notepad file of all the components generated in order to create a component color relation. 3) COMPONENT-COLOR RELATION: Table Code:

sq01

Select Environment and under it select query areas and then double click on standard area

1) Select Other user group

2) Double click on category management


66

Click on Find and enter ZART_EAN

SELECT ZART_EAN and press F8


67

Click here and pull the article text file and the rest of the procedure is same as of the first one.

 ABP Quantity for next four weeks: To get the ABP Quantity for the next 4 weeks the first task is computing the Average Selling Price (ASP) for each merchandise. ASP = One year sales value of a product/Quantity sold Planned Quantity = Sales Forecast for next 4 weeks/ASP 4.25) DESIGNING THE LAYOUT FOR PHASE III:


68

4.26) LOGICS FOR MIS AUTOMATION PHASE III:  TY and LY Season Store Stock Quantity: Arrived at with the help of the stock report and the season master. TY is This Year means SS14 season store stock quantity and LY is Last Year (except) SS-14 Season Store Stock Quantity.  Net Week Cover (TY Season): TY Season store stock Quantity/average of last four week sales quantity.  Net Week Cover (ALL Season): TY + LY Season store stock Quantity/average of last four week sales quantity.  Forward Week Cover (TY Season): TY Season store stock Quantity/average of next four week sales quantity.  Net Week Cover (ALL Season): TY + LY Season store stock Quantity/average of next four week sales quantity.


69

CHAPTER 5

RESULTS


70

5.1) RESULTS FROM THE PHASE I OF MIS AUTOMATION: 5.2)

MEN’s CATEGORY:

As we have seen that for the Men’s Category, Linear Regression method of sales forecasting is used hence with this method predicting the sales for the next four weeks.

1 (17-23 FEB) 2 (24-2 MAR) 3 (3-9MAR) 4 (10-16MAR) 5 (17-23MAR) 6 (24-30MAR) 7 (31-6APR) 8 (7-13APR) 9 (14-20APR) 10(21-27APR)

ACTUAL SALES

Earlier Sales Forecasting

Linear Regression Method of Sales Forecasting

Forecasti ng Error Earlier

Error %age

548.13 546.85 551.58 603.63 660.83 715.42 745.68 794.35 815.26 845.67

735.68 751.64 698.04 833.62 841.42 943.81 951.4 959.35 980.43 899.49

522.52 557.67 592.82 627.97 663.12 698.27 733.42 768.57 803.72 838.87

-187.55 -204.79 -146.46 -229.99 -180.59 -228.39 -205.72 -165 -165.17 -53.82

-34.2 -37.4 -26.6 -38.1 -27.3 -31.9 -27.6 -20.8 -20.3 -6.4

Forecasti Error ng Error %age New 25.6 -10.8 -41.2 -24.3 -2.3 17.2 12.3 25.8 11.5 6.8

4.67 -1.98 -7.48 -4.03 -0.35 2.4 1.64 3.25 1.42 0.80

The above table shows the comparison of sales forecast with the help of Linear Regression method for the week 7 to week 10 with respect to the manual sales forecasting as well as the actual sales that took place. It can be clearly seen from the above table that the sales forecast for the week 7 with the manual sales forecast is showing a percentage error of 27.6% where as with the linear regression method it is showing a percentage error of just 1.64%. Similarly for the week 8, 9 and 10 also the percentage error in case of the linear regression method is much less than that of the manual sales forecast. NOTE: The negative sign in the forecasting %age error shows that the actual sales is less that the predicted sales. 5.3)

Non Apps CATEGORY: ACTUAL SALES

1 (17-23 FEB) 2 (24-2 MAR)

318.81 330.81



337.61 347.38

308.55 353.46

Forecasting Error Error Earlier %age

-18.8 -16.6

-5.9 -5

Forecasti ng Error New

10.3 -22.7


71

Error %age

3.2 -6.8

3 (3-9MAR) 4 (10-16MAR) 5 (17-23MAR) 6 (24-30MAR) 7 (31-6APR) 8 (7-13APR) 9 (14-20APR) 10(21-27APR)

331.53 425.9 492.77 517.18 549.31 597.81 635.34 670.58

238.26 406.28 409.85 458.66 478.35 490.56 512.85 568.69

398.37 443.28 488.19 533.1 578.01 622.92 667.83 712.74

93.3 19.6 82.9 58.5 70.96 107.25 122.49 101.89

28.1 4.6 16.8 11.3 12.9 17.9 19.3 15.2

-66.8 -17.4 4.6 -15.9 -28.7 -25.11 -32.49 -42.16

-20.2 -4.1 0.9 -3.1 -5.2 -4.2 -5.1 -6.3

The above table shows the comparison of sales forecast with the help of Linear Regression method for the week 7 to week 10 with respect to the manual sales forecasting as well as the actual sales that took place. It can be clearly seen from the above table that the sales forecast for the week 7 with the manual sales forecast is showing a percentage error of 12.9% where as with the linear regression method it is showing a percentage error of just 5.2%. Similarly for the week 8, 9 and 10 also the percentage error in case of the linear regression method is much less than that of the manual sales forecast.

5.4)

KIDS CATEGORY: ACTUAL SALES


129.4 128.4 147.78 200 218.41 250.61 279.52 299.98 315.26 346.87



114.07 115.69 167.38 251.77 255.12 288.37 299.31 288.63 298.11 276.38

112.8 139.32 165.84 192.36 218.88 245.4 271.92 298.44 324.96 351.48


15.33 12.71 -19.6 -51.77 -36.71 -37.76 -19.79 11.35 17.15 70.49

11.8 9.9 -13.3 -25.9 -16.8 -15.1 -7.1 3.8 5.4 20.3


16.6 -10.92 -18.06 7.64 -0.47 5.21 7.6 1.54 -9.7 -4.61

Error %age

12.8 -8.5 -12.2 3.8 -0.2 2.1 2.7 0.5 -3.1 -1.3

The above table shows the comparison of sales forecast with the help of Linear Regression method for the week 7 to week 10 with respect to the manual sales forecasting as well as the actual sales that took place.


72

It can be clearly seen from the above table that the sales forecast for the week 7 with the manual sales forecast is showing a percentage error of 7.1% where as with the linear regression method it is showing a percentage error of just 2.7%. Similarly for the week 8, 9 and 10 also the percentage error in case of the linear regression method is much less than that of the manual sales forecast.

5.5)

WOMEN WESTERN CATEGORY: ACTUAL SALES


368.07 378.94 445.81 491.07 587.36 617.24 652.34 701.26 756.79 799.85



436.23 445.21 507.82 579.69 589.25 650.43 683.75 655.01 675.08 626.16

344.53 399.28 454.03 508.78 563.53 618.28 673.03 727.78 782.53 837.28


-68.16 -66.27 -62.01 -88.62 -1.89 -33.19 -31.41 46.25 81.71 173.69

-18.52 -17.49 -13.91 -18.05 -0.32 -5.38 -4.8 6.6 10.8 21.7


23.54 -20.34 -8.22 -17.71 23.83 -1.04 -20.69 -26.52 -25.74 -37.43

Error %age

6.4 -5.37 -1.84 -3.61 4.06 -0.17 -3.2 -3.8 -3.4 -4.7

The above table shows the comparison of sales forecast with the help of Linear Regression method for the week 7 to week 10 with respect to the manual sales forecasting as well as the actual sales that took place. It can be clearly seen from the above table that the sales forecast for the week 7 with the manual sales forecast is showing a percentage error of 4.8% where as with the linear regression method it is showing a percentage error of just 3.2%. Similarly for the week 8, 9 and 10 also the percentage error in case of the linear regression method is much less than that of the manual sales forecast.


73

5.6)

WOMEN ETHNIC CATEGORY: ACTUAL SALES


293.94 320.27 393.67 300.08 359.95 347.19 342.56 372.86 386.47 359.68



344.61 355.01 344.87 347.52 389.78 428.35 410.69 468.98 475.32

293.94 307.105 335.96 338.0067 351.2333 335.74 341.66 342.87 340.09


-24.3 38.7 -44.8 12.4 -42.6 -85.79 -37.83 -82.51 -115.64

-7.6 9.8 -14.9 3.5 -12.3 -25.0 -10.1 -21.3 -32.2


Error %age

26.3 86.6 -35.9 21.9 -4 6.82 31.2 43.6 19.59

8.2 22 -12 6.1 -1.2 2.0 8.4 11.3 5.4

As already discussed earlier that for the women ethnic category the sales forecasting technique that has to be used is moving average method for which we have to take the sales values for the previous three weeks, hence if we have to forecast for the week 7 then we take the sales values from week 4 – 6. But from week 7 onwards we do not have the actual sales values and therefore for predicting the sales for the future weeks we have to take the forecast values. So for week 8 we take the actual sales values of week 5 and 6 and forecast sales values of week 7. Similarly for week 9 we have to take the actual sales value of week 6 and forecast sales values of week 7 and 8. After week 9 we have to consider the forecasted sales values in order to predict the future sales. The above table shows the comparison of sales forecast with the help of Linear Regression method for the week 7 to week 10 with respect to the manual sales forecasting as well as the actual sales that took place. It can be clearly seen from the above table that the sales forecast for the week 7 with the manual sales forecast is showing a percentage error of 4.8% where as with the linear regression method it is showing a percentage error of just 3.2%. Similarly for the week 8, 9 and 10 also the percentage error in case of the linear regression method is much less than that of the manual sales forecast.


74

CHAPTER 6

CONCLUSION


75

6.1)

MIS AUTOMIZATION SALES FORECAST CONCLUSIONS:

My objective through this project was to help in automizing the sales analysis as well as the sales forecasting by building and providing the logics for the MIS. The various mentioned analysis for the sales forecasting were horizontal as well as trend pattern of sales data, where horizontal pattern is that where the sales data keeps on increasing as well as decreasing around a constant mean which in this case was applicable to the women ethnic category. Whereas the trend pattern is that where there is a constant rise in the sales data over a period of time for which a linear regression method of sales forecast is applicable and in my project it is applied to the Men’s, Kids, Non-Apps as well as the Women Western Category. This project has shown that with these methods of sales forecasting the percentage error from the actual sales has decreased leading to an increase in the forecast accuracy. An accurate sales forecasting would lead to the following conclusions:

 Increased Turnover: Merchandise that customers want is more readily available at times when they want to make purchases.

 Reduced Amounts of Markdowns: Because merchandise purchases in relation to planned sales and stock levels are anticipated, there is less likelihood of being in an overbought position and having to make markdowns.  Improved Ability to Maintain Markups: As the stock in hand would not be in much larger quantity hence the maintained markups would sustain and there will be no need in giving markdowns.

 Maximized Profits: A balanced assortment of merchandise leads to more sales and an increase in profits because items will not remain in stock for too long and would be difficult to sell. Greater profits can result because the buyer is informed about both fast-selling items that should be reordered quickly and slow-selling items that should be dropped.

 Minimized Inventory Investments: An accurate merchandise plan helps to determine how much money should be spent on merchandise. Ideally, a planner makes the smallest investment possible in goods that will satisfy customer demands and sell well enough to build store profits.


76

CHAPTER 7

REFRENCES


77

7.1)

BOOKS:

 (Kress, G., 1985, Practical techniques of business forecasting, Westport)  (Business Forecasting with Student CD [J. Holton Wilson, Barry Keating, Tata McGrawHill)  Forecasting systems for operations management: Stephen A. Delurgio and Carl D.Bhame, 1991, (Business One Irwin, Homewood, IL)  Forecasting Methods for Management by Spyros Makridakis and Steven C. Wheelwright (1977, Hardcover)  Sales Forecasting Management: A Demand Management Approach By John T. Mentzer & Mark A. Moon)

7.2)

ARTICLES AND JOURNALS:

 http://www.forecastingprinciples.coM  http://fearp.usp.br/marketing/artigos/  http://faculty.philau.edu/frankc/ntc/s01-ph10  http://arxiv.org/ftp/arxiv/papers/1303/1303.0117

7.3)

WEBSITES:

 http://sbinfocanada.about.com/od/cashflowmgt/a/salesforecast.htm  http://smallbusiness.chron.com/methods-techniques-sales-forecasting-4693.html  http://managementinnovations.wordpress.com/2008/12/11/methods-of-salesforecasting/  http://blog.getbase.com/5-essential-sales-forecasting-techniques


78