USING ACTIVITY BASED COSTING AND THEORY OF CONSTRAINTS TO ENHANCE

DECISION MAKING AT DUOGRAPHICS B.V.: A CASE STUDY

Dr. Philip G.M.C. Vergauwen

Faculty of Economics & Business Administration
Department of Accounting & Information Management
Universiteit Maastricht
e-mail: p.vergauwen@aim.unimaas.nl

Drs. Christian C.J.M.C. Kerckhoffs

Faculty of Economics & Business Administration
Department of Quantitative Economics
Universiteit Maastricht
e-mail: c.kerckhoffs@ke.unimaas.nl

Revised for resubmission

Issues in Accounting Education
July 2005

(CASE TEXT FOR STUDENTS)

Abstract

This case is about Activity Based Costing (ABC) and throughput accounting (TA) as accounting

tools to “structure” technical (process) insights in an accounting context. The case shows how

working-floor insights and production process data can be used in the computation of income

statements that are relevant managerial decision making. The throughput accounting part of the

case stimulates a discussion on (short run) avoidable costs and hence provides necessary input for

short-run managerial decision making.

It is essential to understand that ABC is a method that tries to “frame” data in a managerial

decision making context. Models, figures and theories only have value to the extent that they help

us to understand and explain what happens in the company. ABC is an allocation method cum

story: it is an arbitrary way to allocate costs – as arbitrary as any other allocation method – but it

helps in communicating how the (production) process works and, therefore, tries to upgrade the

value of accounting in the decision making process. If, e.g., a good driver is found and if there is

1
statistical evidence that supports the choice of and/or work-floor feeling on such a driver, using

this cost driver in the calculations of product (line) profitability makes the accounting much more

“vivid” and important to decision making.

Mr. Van Schenkel, founder and owner of small company producing colour printed T-shirts, has

some ideas on why multi-coloured T-shirts are more expensive in production than T-shirts with

fewer colours printed on them. Process mapping and accounting data are available to set up an

ABC system that will allow to test the validity of Mr. Van Schenkel’s ideas and to come to “better”

accounting calculations of product profitability. The case illustrates how ABC calculations can

support managerial decision making and lead to ABM (activity based management), enabling the

company to better understand how its process works and how this understanding can be put to use

in an accounting context (income statement, performance evaluation, product profitability, …).

Introduction: Company and Industry Background

DuoGraphics B.V. is a (non existent) custom screen printing business whose major customers are

local high schools, youth organizations, sporting clubs, and small businesses. The company

produces a variety of T-shirts, silk screened in as many as six different colours. Normally,

customers provide the blank T-shirts. If not, the company buys blank T-shirts and just charges the

customers for them, without a mark-up. Thus, the only “raw material” used by the company is the

ink. Table 1 shows the ink used by each of the products. The raw material (ink) is to be treated as a

direct cost. This means that the cost is incurred only as the product is actually produced (direct

material, i.e. ink costs amounted to €3,420).

2
There are virtually no entry barriers to the screen printing business: start-up costs are small and the

printing technique is one that is learned easily. Consequently, the business is highly competitive

and companies try to innovate continuously.

<INSERT TABLE 1 ABOUT HERE>

In its fourth year of operations, DuoGraphics B.V. wants to start with activity-based costing (ABC)

and ABC-management (ABCM) and would like to integrate this cost system with insights from the

theory of constraints (TOC). A first step in this implementation process entailed an analysis of the

activities that are necessary in the custom screen printing business. The company’s founding father,

owner and chief executive, Mr. van Schenkel, has identified the following main activities: (i)

production, (ii) machining, (iii) set-ups, (iv) quality-control, (v) reclaiming, (vi) purchasing and

(vii) sales and marketing. Mr. van Schenkel has also analysed the income statement for the year

ended in order to reclassify the expenses (totalling €129,652) from the traditional account

classification to an activity-based costing (ABC) classification. Tracing to activities was

accomplished through interviews and an analysis of the accounting records together with all types

of supporting documents. The following breakdown was the result of this tracing to activities and is

shown in table 2:

<INSERT TABLE 2 ABOUT HERE>

The set-up, quality control, and reclaiming activities had no costs directly traced to them because

analysis revealed that they used resources from the production labour and – to a much lesser extent

– from the machines activities categories.

As ABCM requires the identification of cost drivers for each of the activities, Mr. van Schenkel has

also gathered all possible and relevant information for “process mapping”. Process mapping means

relating all activities with cost drivers in such a way that it becomes clear why and where resources

3
such as machine and labour hours are consumed in the process of producing, marketing and selling

the company’s products. In other words, a process map shows the relationship of each activity to

other activities and to the products and is the key instrument to determine product costs, cost

drivers and capacity of each activity. Table 3 provides you with the monthly data about the

resources consumed in the production process (weekly data are given in appendix 1). It shows the

result of this mapping process and indicates, where applicable, the capacity of each activity in

terms of the selected cost driver.

<INSERT TABLE 3 ABOUT HERE>

Mr. van Schenkel now wants to determine the cost, cost drivers (output measures), and capacity of

each activity, using the information available to him. The company's products are identified

according to the number of colours printed on the shirt. Because the maximum number of colours

printed was six, there are six different products, to be referred to as “clr1” (a single colour T-shirt)

up to “clr6” (a T-shirt with six colours printed on it). The associated EXCEL dataset contains two

worksheets with (i) weekly data for the resources consumed for six out of the seven activities listed

above together with the corresponding weekly outputs of the six different types of shirts and (ii)

monthly “purchasing orders” data. The data concerns the last year of operations. The names of the

variables in the dataset should be self-evident.

The screen printing of six-colour T-shirts is significantly more complex than for one- and two-

colour T-shirts. Thus one would a priori expect e.g. a six-colour T-shirt to require more machine

time, more reclaiming time, more set-up time, more quality control, and more production labour. In

addition, more purchase orders (e.g. for ink) are required for six-colour T-shirts (see also table 1).

Linking up with this, ABC cost calculations require the calculation of “cost driver factors” showing

the extent to which each product makes use of the various activities (i.e. the associated cost drivers)

in the production process. This means that, for all seven cost drivers in our dataset, we can use

multiple regression to estimate the “cost driver factors” from our production data.

4
Concerning these cost drivers, the annual cost of the activity to which the driver is linked can be

calculated. For example, consider the machining activity. The annual cost for machines is €18,200

per year (see table 2). The annual machine capacity can be calculated as follows: 40 hours/week

and 52 weeks per year makes up 2,080 “normal” machine hours per year. Mr. Van Schenkel

suggests that you use the available data to estimate the machine hours required for each (type of) T-

shirt. In fact, mr. Van Schenkel has always been - and still is - in the habit of getting his hands dirty

himself. As a consequence, he has an excellent understanding of the day-to-day operations of his

establishment. He is convinced that the amount of machining does not simply depend on the total

number of shirts produced, but also on the number of colours printed on these shirts. More

precisely, his point is as follows: “Making e.g. a 2-colour shirt means that you have to pull the

shirts through the machine two times, once for each different colour. Logically, this should imply

that to produce a given quantity of such shirts, you need twice as much machine time as for 1-

colour shirts. Similarly for 3-, 4-, 5- and 6-colour shirts, of course!”1. A similar story holds for the

production labour activity. It can be explained and traced to the products in the same way as the

machine activity. The total labour cost amounts to €68,229 per year, the output measure or cost

driver is the number of labour hours, and the annual capacity of fixed labour hours per year is 5,760

labour hours (if calculated on the basis of 3 fte each working 48 weeks per year and 40 hours per

week). Mr. van Schenkel wants you to find out whether (as was the case with machine hours) total

production labour not only depends on the number of shirts produced, but also on the number of

colours printed on a shirt2.

1
Therefore, he wants you to test the following machine cost calculation formula: MACH(x-clr) = a*x, where a is the

parameter to be estimated.

MACH(x-clr) stands for the machine hours needed to produce a single T-shirt with x colours printed on it.

E.g. MACH(1-clr) = the amount of machine hours needed for the production of one single-colour T-shirt.

5
Mr. Van Schenkel also asks you to find out whether it would not be better to “split-up” (total)

production labour into set-up labour, quality control labour, reclaiming labour and “pure”

production labour (i.e. the remainder). Concerning set-up labour, Mr. Van Schenkel reasons as

follows: “To set up the machine requires some basic operations which are always needed, no

matter how many colours we want to print. After that, each additional colour requires the same

operations (e.g. putting in the right ink colour), and should therefore take an equal amount of

labour hours. E.g. if the ‘basic operations’ require a setup lasting for 0.6 minutes (i.e. 0.01 hours),

and each different colour requires an additional 0.3 minute setup, I would expect a ‘cost driver

factor’ pattern like 0.015, 0.020, 0.025, 0.030, 0.035 and 0.040 for 1- up to 6-colour shirts

successively.” This would suggest a formula similar to the one for total labour hours discussed

above. Mr. van Schenkel does, however, point out to you that the formula discussed above might

not hold for the other three types of labour driven activity (i.e. quality control, reclaiming and

“pure” production labour).

Reclaiming, set-up, and quality control activities are performed using production labour. Therefore,

these activities do not have their own annual cost and capacity. They have no cost of their own, and

the capacity is limited by the capacity of the production labour activity. In the case of the machine

and the production labour activity, the capacity represents the number of hours that can reasonably

be expected to be available, given the estimated cost for the activity.

The specification of the capacities of the administrative activities, i.e. purchasing and sales and

marketing, is more difficult. The line of reasoning is usually as follows. Last year, €28,566 was

spent on the sales and marketing activity and 2,562 sales orders were processed during that year.

Hence, it would be difficult to argue that the capacity was less than these 2,562 sales orders. On the

other hand, it might be possible to argue that the capacity is much greater because there was unused

2
In fact, he suggests to test the following, slightly more complicated formula: LH(x-clr) = a*x + b, where a and b are

the parameters to be estimated (note that the machine cost formula is actually a special case, if we choose b=0).

6
capacity last year. To use last year's experience as the capacity is a conservative approach. Mr. van

Schenkel suggests that you use purchase orders and sales orders as “cost drivers” for the purchasing

and sales & marketing overhead costs.

Concerning the sales and marketing activity, he argues as follows: “It’s clear that making e.g. a 6-

colour shirt requires more machine time than a 1-colour shirt, because you have to pull the shirts

through the machine six times (once for each different colour). However, it’s not at all clear why a

6-colour shirt would require more sales and marketing time than a 1-colour shirt3.”

Mr. van Schenkel has worked out a basic regression model for the sales & marketing activity. It

looks as follows:

(C) SOt = β 0 + β1clr1t + β 2 clr 2 t + β3 clr3t + β 4 clr 4 t + β5 clr 5 t + β6 clr 6 t + εt t = 1,…,48

The βx-symbols (x=1,…,6), which have the same meaning as our previous SO(x-clr), are standard

notation in statistics. This model (C), the “complete model”, is an unrestricted model, meaning that

the coefficients can be freely estimated, i.e. they are not thought to obey a particular pattern. The

EXCEL-results of the “complete model” regression are reported in Table 44.

<INSERT TABLE 4 ABOUT HERE>

By means of first impression, Mr. van Schenkel notes that the results look excellent: all the

independent variables (i.e. types of shirts) are significant at the 5% level (the highest p-value is

0.042), and we have a nice R2 of 83.2%. Upon closer inspection, two things strike him, however:

1) The intercept of the model is negative, but very insignificant. this raises the obvious question

whether a model like this ought to contain an intercept at all.

3
This would lead to the following formula: SO(x-clr) = b, where b is the parameters to be estimated (note that this is

again a special case of the formula for production labour, if we choose a=0).
4
Using the EXCEL data appendix, this output can be obtained as follows: select the menu Tools > Data Analysis,

choose the “Regression” option, select input Y-range H1:H49, input X-range B1:G49, tick “labels” and click “OK”.

7
2) Mr. van Schenkel also notes that the six “slope” coefficients of the model are not too far apart:

in fact, they all seem to hover around 0.045. This obviously links up with his prior hunch, and it

raises the issue whether those coefficients might actually be the same, as expressed by the

formula SO(x-clr) = b, which can be converted into a null hypothesis in terms of the coefficients

of model (C).

Activity-Based Cost Management Model & Theory of Constraints

The information in tables 2 and 3 (see also the data in appendix 1) is to be used to calculate the

charging rates for each of the activities for which there are annual costs and annual capacities. The

charging rates are the ratio (annual) cost/capacity for each activity. Together with the information

given in tables 5 and 6, these charging rates can then be used to create an income statement

showing product profitability based on the extent to which each product makes use of the various

activities in the process.

<INSERT TABLES 5 AND 6 HERE>

Goldratt (1992) introduces three operational measures - throughput, operating expense, and

inventory - to provide feedback to operations management as they seek to make operating decisions

that will enhance profits, cash flow, return on sales and residual income.

Throughput can be defined as the revenue minus the cost of the raw materials based on the number

of units sold (not produced). The Goldratt (1992) definition of operating expense comprises the

expenses that are incurred to convert raw materials to throughput. Thus, he regards all expenses

other than raw materials as operating expenses. Inventory represents the cost of all the raw

materials that have not yet been sold plus the cost of the more traditional assets that are used in the

business (plant and equipment).

8
The foregoing is a simplified version of Goldratt's definition: you could argue that more than just

raw materials are to be deducted from revenues to arrive at throughput. So, you could deduct any

resource that is purchased on an as-needed basis and not acquired in lump sum in advance of its

use. Thus, in Goldratt's view, labour is not treated like raw materials. Instead, labour is regarded as

an operating expense because a company typically purchases labour services in a lump sum, and, if

the services are not used, the company has paid for the services anyway.

In Goldratt's view, a process is a series of dependent events (activities), and the goal is the

maximization of the process throughput. The primary challenge to maximizing the throughput is in

identifying the activity or activities that are the constraints to throughput. Once the constraining

activity or activities have been identified, attempts to increase the throughput without increasing

operating expense or increasing inventory can be made.

Requirements

(A) Compute a regular Income Statement for the year ended showing the per-product throughput,

return-on-sales and try to define a metric per product that gets as close as possible to a return-

on-investment (ROI) and residual income (RI) metric. (Note: assume a capital charge of 8% p.a.

for the RI metric and allocate costs on the basis of production volume (traditional allocation).

Assess your findings and discuss the pricing policy and product portfolio of DuoGraphics B.V.

(B) Comment on Mr. Van Schenkel’s regression analysis output (model (C)) with respect to the

sales & marketing activity and discuss the two findings that strike him. Is there an alternative

model you would like to test that satisfies Mr. Van Schenkel’s hunch? How would you compare

such an alternative model with the complete (unrestricted) model (C)?

(C) Mr. Van Schenkel asks you to run regressions for all other activities and asks you to “test his

ideas” as to the what really drives the costs of the activities left to analyse.

9
(D) Use the “process mapping” information to set up an ABC-system that allows you to compute

“charging rates” for each driver and “cost driver factors” per product per activity. Also an

“ABC Income Statement” showing the per-product throughput, return-on-sales, “return-on-

investment” and “residual income” on the basis of your ABC-system. Finally, try to calculate

the “cost of unused capacity” (see Cooper & Kaplan (1992)) and assess your findings. Will

your findings lead to a different pricing policy and product portfolio decision at DuoGraphics

B.V.?

10
TABLES

11
 TABLE 1
DuoGraphics B.V.: Ink used in Screen Printing

Product Millilitres per unit product (average)

One-colour shirt 1,0
Two-colour 1,0
Three-colour 1,0
Four-colour 1,5
Five-colour 2,0
Six-colour 2,0

Note: the cost of the ink is € 50,00 per liter.

12
 TABLE 2
DuoGraphics B.V.: Tracing Expenses to Activities

Activity
Production labour € 68.229
Machines € 18.200
Purchasing € 14.657
Sales and marketing € 28.566
Total Costs €129.652

13
 TABLE 3
Duographics B.V.: Process Mapping Information (Monthly Data)5

Reclaiming

Purchasing
Production

production
Machining

Marketing
Sales &
Quality
Activity:

Set-ups

“Pure”
control

Sales orders
Reclaiming

Purchasing
production
Cost Machine

Quality
Labour

“Pure”
control
labour

labour

labour

labour
Set-up

orders
driver:
hours

hours

hours

hours

hours

hours
Month LH6 MACH SULH QCLH RECLLH PRODLH7 PO SO
Jan 407.50 87.75 67.75 124.75 99.75 115.25 47 146
Feb 385.50 90.25 59.75 121.25 97.25 107.25 47 124
Mar 514.25 107.25 92.00 128.75 137.25 156.25 58 206
Apr 652.75 138.75 103.75 149.00 167.25 232.75 66 235
May 638.00 153.00 115.75 137.50 181.25 203.50 75 238
Jun 647.50 151.00 117.25 131.00 188.50 210.75 80 248
Jul 629.00 148.75 108.25 134.50 179.75 206.50 75 250
Aug 607.75 144.00 103.50 136.75 172.75 194.75 72 227
Sep 584.75 130.50 92.50 148.25 161.00 183.00 67 240
Oct 542.25 122.00 87.00 141.25 148.25 165.75 64 204
Nov 460.50 101.75 74.75 131.50 118.50 135.75 48 171
Dec 652.00 157.40 122.00 127.25 173.50 229.25 89 273
Total 6,721.75 1,532.40 1,144.25 1,611.75 1,825.00 2,140.75 788 2562

5
For weekly data, see appendix 1.
6
Note that 3 full-time equivalents are hired on a fixed basis and used for production. Extra production labour

(excluding the 40 hours per week per fte) is hired on a temporary basis.
7
PRODLH is total production labour minus the set-up, quality control and reclaiming labour hours.

14
 TABLE 4
Duographics B.V.: Sales & Marketing – Regression analysis output

Model (C) Sales & marketing

Regression Statistics
Multiple R 0.912
R Square 0.832
Adjusted R Square 0.808
Standard Error 5.308
Observations 48

ANOVA
df SS MS F Significance F
Regression 6 5737.991 956.332 33.940 0.000
Residual 41 1155.259 28.177
Total 47 6893.250

Coefficients Standard Error t Stat P-value

Intercept -0.5029 4.007 -0.126 0.901
CLR1WK 0.0425 0.004 9.678 0.000
CLR2WK 0.0479 0.007 6.674 0.000
CLR3WK 0.0372 0.010 3.894 0.000
CLR4WK 0.0400 0.017 2.403 0.021
CLR5WK 0.0514 0.025 2.099 0.042
CLR6WK 0.0452 0.011 3.934 0.000

15
 TABLE 5
DuoGraphics B.V.: Sales

Product Sales units Sales (€)

1-clr 36.000 €90.000
2-clr 9.000 €29.250
3-clr 5.900 €19.500
4-clr 2.400 €7.800
5-clr 1.800 €5.850
6-clr 4.800 €15.600
Total 59.900 €168.000

Note: These figures reflect actual sales for the past year.

16
 TABLE 6
Duographics B.V.: Sales (Monthly Data)8

Month CLR1MON CLR2MON CLR3MON CLR4MON CLR5MON CLR6MON TOTAL

Jan 2,250 510 315 130 105 225 3,535
Feb 1,900 495 355 135 100 260 3,245
Mar 2,800 700 440 180 125 350 4,595
Apr 3,050 815 470 230 150 425 5,140
May 3,350 915 580 265 185 470 5,765
Jun 3,450 855 600 255 205 500 5,865
Jul 3,580 925 590 120 195 485 5,895
Aug 3,300 790 555 250 160 440 5,495
Sep 3,240 785 495 210 155 395 5,280
Oct 2,850 685 480 195 150 380 4,740
Nov 2,480 615 385 165 105 295 4,045
Dec 3,750 910 635 265 165 575 6,300
Total 36,000 9,000 5,900 2,400 1,800 4,800 59,900

Note: These figures reflect actual monthly sales units for the past year, as provided in the
associated EXCEL-datafile, “monthly data” tab.

8
For weekly data, see appendix 2 or the “weekly data” tab of the associated EXCEL-datafile.

17
RELATED LITERATURE

Baxendale, S. J., and M. Gupta. 1997. Integrating TOC and ABCM in a Healthcare Company.
Journal of Cost Management (11).
Baxendale, S. J., and M. Gupta. 1998. Aligning TOC & ABC for Silkscreen Printing. Management
Accounting (April): 39-44.
Campbell, R. 1997. Designing an Information System using ABC and TOC, Journal of Cost
Management (11).
Cooper, R., and R. S. Kaplan. 1992. Activity-Based Systems: Measuring the Costs of Resource
Usage. Accounting Horizons (September): 1-13.
Cooper, R., and R. Slagmulder. 1999. Strategic Cost Management: Integrating ABC and TOC.
Management Accounting (February): 20-21.
Demmy, S., and J. Talbot. 1998. Improve Internal Reporting with Activity-based Costing and
Theory of Constraints. Management Accounting (November): 18-24.
Goldratt, E.M., and J. Cox. 1992. The Goal: A Process of Ongoing Improvement, 2nd (revised)
edition, Croton-on-Hudson, NY: North River Press.
Goldratt, E.M., and R. Fox. 1989. The Importance of a Systems Constraint. The Theory of
Constraints Journal 1 (February): 5-7.
Holmen, J. 1995. ABC vs. TOC: It’s a Matter of Time. Management Accounting (January): 37-40
Noreen, E., D. Smith, and J.T. Mackey. 1995. The Theory of Constraints and Its Implications for
Management Accounting. Great Barrington, MA: The North River Press.

18
 APPENDIX 1
DuoGraphics B.V.: Weekly Data Cost Drivers (see EXCEL file, “weekly data” tab)

Month Week LH MACH SULH QCLH RECLLH PRODLH PO9 SM

Jan 1 96.00 17.75 18.25 29.75 19.25 28.75 47 39
2 84.50 17.25 12.50 32.75 16.75 22.50 34
3 93.50 17.25 16.00 33.50 19.00 25.00 33
4 133.50 35.50 21.00 28.75 44.75 39.00 40
Feb 1 88.50 20.25 12.25 29.50 20.75 26.00 47 29
2 112.00 25.25 16.50 36.00 27.50 32.00 37
3 92.50 28.00 15.75 20.50 29.75 26.50 34
4 92.50 16.75 15.25 35.25 19.25 22.75 24
Mar 1 130.75 27.00 21.50 37.00 35.25 37.00 58 54
2 141.25 32.50 26.25 35.50 40.25 39.25 49
3 126.25 27.00 24.75 24.50 35.00 42.00 58
4 116.00 20.75 19.50 31.75 26.75 38.00 45
Apr 1 177.75 41.75 29.75 38.50 54.50 55.00 66 61
2 179.25 31.25 25.75 37.25 32.25 84.00 60
3 141.75 29.50 24.00 31.50 39.75 46.50 57
4 154.00 36.25 24.25 41.75 40.75 47.25 57
May 1 188.25 41.75 31.75 39.75 53.75 63.00 75 62
2 166.00 46.50 32.75 25.00 53.00 55.25 68
3 152.50 33.50 26.25 41.50 38.75 46.00 56
4 131.25 31.25 25.00 31.25 35.75 39.25 52
Jun 1 169.25 42.50 30.00 33.50 54.25 51.50 80 65
2 170.00 43.00 30.75 36.75 51.75 50.75 65
3 148.00 31.25 26.00 25.75 43.25 53.00 56
4 160.25 34.25 30.50 35.00 39.25 55.50 62
Jul 1 154.50 33.75 26.00 34.50 35.75 58.25 75 60
2 134.00 29.50 22.50 30.25 36.50 44.75 64
3 211.75 61.50 37.25 33.75 79.25 61.50 62
4 128.75 24.00 22.50 36.00 28.25 42.00 64
Aug 1 156.25 32.00 25.25 30.75 46.25 54.00 72 63
2 132.25 29.00 25.50 31.00 29.50 46.25 60
3 154.50 36.00 26.50 37.75 42.00 48.25 55
4 164.75 47.00 26.25 37.25 55.00 46.25 49
Sep 1 169.50 49.50 29.25 33.75 57.50 49.00 67 64
2 150.00 34.00 21.75 43.75 45.00 39.50 44
3 135.75 26.25 21.50 40.00 28.50 45.75 66
4 129.50 20.75 20.00 30.75 30.00 48.75 66
Oct 1 151.25 33.75 23.75 36.00 46.50 45.00 64 50
2 121.25 23.50 16.00 43.00 25.50 36.75 53
3 136.75 31.75 25.75 29.50 39.50 42.00 50
4 133.00 33.00 21.50 32.75 36.75 42.00 51
Nov 1 102.50 29.75 14.75 26.25 31.25 30.25 48 36
2 143.00 28.50 23.75 48.50 26.75 44.00 57
3 114.00 22.50 18.00 35.00 30.00 31.00 38
4 101.00 21.00 18.25 21.75 30.50 30.50 40
Dec 1 153.00 35.25 27.25 33.75 40.75 51.25 89 69
2 165.25 42.15 33.75 26.75 47.75 57.00 62
3 168.00 40.00 28.00 28.00 52.75 59.25 73
4 165.75 40.00 33.00 38.75 32.25 61.75 69

9
Only monthly data are available.

19
 APPENDIX 2
DuoGraphics B.V.: Weekly Sales Data X-colour Shirts (see EXCEL file, “weekly data” tab)

Month Week CLR1WK CLR2WK CLR3WK CLR4WK CLR5WK CLR6WK

Jan 1 650 210 125 0 0 0
2 600 90 150 0 0 0
3 600 110 40 30 0 25
4 400 100 0 100 105 200
Feb 1 525 115 0 75 0 30
2 550 140 0 60 0 100
3 400 65 180 0 50 130
4 425 175 175 0 50 0
Mar 1 650 130 200 40 25 100
2 675 110 150 40 100 100
3 750 210 90 20 0 125
4 725 250 0 80 0 25
Apr 1 750 190 0 30 150 200
2 775 195 275 0 0 0
3 800 190 125 25 0 150
4 725 240 70 175 0 75
May 1 1,000 270 0 0 0 275
2 750 260 125 125 100 125
3 825 250 175 125 50 0
4 775 135 280 15 35 70
Jun 1 775 110 200 150 100 125
2 750 140 200 105 105 125
3 950 220 150 0 0 150
4 975 385 50 0 0 100
Jul 1 1,250 400 50 0 0 0
2 750 300 140 0 45 85
3 550 35 280 120 150 400
4 1,030 190 120 0 0 0
Aug 1 975 175 0 75 60 115
2 950 360 105 0 0 0
3 750 235 175 0 0 175
4 625 20 275 175 100 150
Sep 1 650 0 250 100 75 250
2 675 15 175 110 80 125
3 810 450 50 0 0 20
4 1,105 320 20 0 0 0
Oct 1 850 0 120 95 75 125
2 700 290 125 0 0 25
3 725 120 120 0 50 150
4 575 275 115 100 25 80
Nov 1 600 0 125 0 75 100
2 575 600 100 0 0 25
3 650 15 75 80 30 75
4 655 0 85 85 0 95
Dec 1 725 360 250 0 0 125
2 800 400 250 35 0 150
3 1,100 150 135 30 65 150
4 1,125 0 0 200 100 150

20