Chapter 7 Shapiro: Supply Chain Databases Ronald … 7 Shapiro: Supply Chain Databases 17/2/2009 SCM 1 Ronald Batenburg. ... (Kraljic-matrix) ... SCM 2009.ppt [Compatibiliteitsmodus]

MBI course Logistics

Chapter 7 Shapiro:

Supply Chain Databases

17/2/2009 SCM 1

Ronald Batenburg

Schedule

� Chapter 7– Some case examples

– A framework of supply chain decision databases

17/2/2009 SCM 2

“(…) In most instances, more than 80% of the data in transactional databases are irrelevant to decision making. Data aggregations and other analyses are needed to transform the remaining 20% or less, into useful information in the supply chain decision database (…)”

Case example the home/furniture sector

Independent

Stores

Wholesales

Suppliers

Wholesales

Suppliers

Wholesales

Suppliers

E-Procurement organization

Independent

StoresIndependent

StoresIndependentStores

Wholesales

Suppliers

19/2/2007 Logistics 3

Wholesales

Suppliers

Wholesales

Suppliers

Web Order

Database

Management

Web Catalog

Database

Management

Franchise

StoresFranchise

StoresFranchise

StoresFranchiseStores

Suppliers

Wholesales

SuppliersWholesales

Suppliers

GS1 / DAS

Furniture

Product

database

Starting Point:The organization as a supply chain model

SUPPLIERS CUSTOMERSFACILITIES

17/2/2009 SCM 4

Raw materials Finished

Intermediate

productsProcesses

Intermediate

products

Model assumptions:

• no mutual exchange

• no disintermediation

• no external markets

• no time dependencies

Conditions for modeling organizations as asupply chain model

� Model assumptions:– no mutual exchange– no disintermediation– no external markets– no time dependencies

17/2/2009 SCM 5

� Four key Data Elements for decision DB:– Products– Place– Price– Policy

A framework for supply chain data and decisions

Suppliers Facilities Customers

A. Products

17/2/2009 SCM 6

B. Place

C. Price

D. Policy

A. Product data

� Customer– Product aggregation, categories

– Customer value

– BCG-matrix

� Facility

17/2/2009 SCM 7

� Facility– Product aggregation, categories

– Processes, product flows

� Supplier– Product aggregation, categories

– Kraljic matrix

Cf. The socio-technical approach

Production structure Governance structureMacro

Customer order

Segmentation

17/2/2009 SCM 8

Micro

implies

parallel product

streams

Product aggregation bycustomer value portfolio:

calc the Gini coefficient / Lorentz Curve

17/2/2009 SCM 9

Decomposition of Customer Value

Increase the Number ofCustomer

Increase

Customer Gross Profit

Decrease

Cost per Customer

Increase

RelationshipDuration

17/2/2009 SCM 10

$$ Customer Value $$

Decisions based on customer value portfolio management

Market share

High Low

19/2/2007 Logistics 11

Market

growth

High ‘Star’ ‘Wild cat’

Low ‘Cash cow’ ‘Dog’

Decisions based on supplier value portfolio management

Supply risk

High Low

19/2/2007 Logistics 12

Financial importance

High ‘Strategic ‘Lever’

Low ‘Bottle neck’ ‘Routine’

(Kraljic-matrix)

Class Exercise:What Product Data

Aggregation/Categorization applies to Utrecht University?

Customer Facility Supplier

17/2/2009 SCM 13

Strategic ?? ?? ??

Tactical ?? ?? ??

B. Place data

� Customer– Outbound transportation networks

– Location decision

� Facility

17/2/2009 SCM 14

� Facility– Shop floor design

– Location decision

� Supplier– Inbound transportation networks

– Location decision

Location costs are transportation costs - generic determinants

� Shape, weight or size (SKU)

� Value

� Time critical (asap, JIT)

17/2/2009 SCM 15

� Temperature/conservation critical

� Combinatory, sequential conflicts

� Transport-processes combinations

� …

Saving facilities-supply chain costs by shaping the shop floor

1 2 3 4

5

17/2/2009 SCM 16

6

78910

Saving supply chain costs by location optimization: the case of Yamaha

Spare parts factory

Yamaha Motor Europe

Spare parts factory

Yamaha Motor Europe

Before After

delivery deliveryordering ordering

Before After

19/2/2007 Logistics 17

25 Distributors

6000 Dealers

Customers

25 Distributors

6000 Dealers

Customers

C. Price data

� Customer

– Market prices

– Margins

� Facility

17/2/2009 SCM 18

� Facility

– Internal transfer prices

– Direct/indirect costs, ABC

� Supplier

– External transfer prices

Generic Cost Decomposition

Direct

costs

Product

costs

process

costs

17/2/2009 SCM 19

Indirect

costs

Facility

resources

Facility

overhead

costsCosts =

Units * Price/Unit

Example: Gasoline Prices decomposed

Sales price (€/liter)

17/2/2009 SCM 20

Oil price (€/liter)

Generic Cost Relationships

(“gradation”)

costs

costs

unitscosts

17/2/2009 SCM 21

Costs = Units * Price/Unit

Costs =

Units * Price_a/Unit for Units < x

Units * Price_b/Unit for Units > x

(“gradation”)

unitsx

unitscosts

units

Direct facility costs

Process Discrete parts

manufacturing

Packaging Distribution

centers

Units Continuous Discrete Discrete Discrete

Cost driver Raw material

price

Raw material

price and labor

Labor Labor

17/2/2009 SCM 22

price price and labor

Cost

relationship

Linear, simple Linear,

complex

Linear,

complex

Complex

•Very complex: indirect costs – are not related to resource units, but to activities/objectives

•Method: Activity Based Costing

Example: Transfer prices Tasty Chips Supply Chain (Shapiro p. 281-286)

Iowa Cincinatti

Nashville

Chicago

Cleveland

41 M

arkets/cities

17/2/2009 SCM 23

Kanses

Maine

Texerkana

Peoria

Farm

CooperativesPlants

Kanses City

Louisville

Little Rock

41 M

arkets/cities

DCs

Conclusions from the Tasty Chips Supply Chain example

� FROM SUPPLIER TO PLANT– Purchase prices differ between corn and potatoes (corn is more expensive)

– Purchase prices differ between supplier-product combination (Iowa have potato shortages)

– Transport prices differ between corn and potatoes (corn is more expensive)

– Transport prices differ between supplier-plant combination (Maine is further away)

17/2/2009 SCM 24

– Transport prices differ between supplier-plant combination (Maine is further away)

� FROM PLANT TO PLANT– Unit change prices differ between corn and potatoes (potato is more expensive)

– Baking and packaging prices differ between corn and potatoes (corn is more expensive)

– Baking and packaging prices differ between factories (Peoria and Texarkana are ‘inefficient’)

� FROM PLAN TO DC AND MARKET– No substantial differences between product, location or production-location prices

D. Policy data

� Customer

– Demand forecasting

� Facility

– Production planning

17/2/2009 SCM 25

– Production planning

� Supplier

– Procurement

Policy data - customer:how to forecast/predict demands

annual beer consumption

86

88

90

annual beer consumption

?

17/2/2009 SCM 26

74

76

78

80

82

84

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

?

Demand Forecasting Methods

Quantitative

Forecasting

Causal

Models

Time Series

Models

17/2/2009 SCM 27

Linear

Regression

Models

ExponentialSmoothing

MovingAverage

Models

Trend

Projection

TimeTimeResponseResponse

YYii

Moving TotalMoving Total((nn = 3)= 3)

MovingMovingAvg. (Avg. (nn = 3)= 3)

19931993 44 NANA NANA

Moving Average Solution

MAMAnn

nn==∑∑ Demand in Demand in Previous Previous PeriodsPeriods

17/2/2009 SCM 28

19941994 66 NANA NANA

19951995 55 NANA NANA

19961996 33 4 + 6 + 5 = 154 + 6 + 5 = 15 15/3 = 5.015/3 = 5.0

19971997 77 6 + 5 + 3 = 146 + 5 + 3 = 14 14/3 = 4.714/3 = 4.7

19981998 NANA 5 + 3 + 7 = 155 + 3 + 7 = 15 15/3 = 5.015/3 = 5.0

Moving averages

84

86

88

90

annual beer consumption MA (n=4) MA (n=3) MA (n=2) MA (n=1)

17/2/2009 SCM 29

74

76

78

80

82

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Ft= F

t-1 + a· (At-1 - F

t-1)

Time ActualForecast, Ft

(a = .10)

1993 180 175.00 (Given)

Exponential Smoothing Solution

17/2/2009 SCM 30

1993 180 175.00 (Given)

19941994 168168 175.00 + .10(180 175.00 + .10(180 -- 175.00) = 175.50175.00) = 175.50

19951995 159159 175.50 + .10(168 175.50 + .10(168 -- 175.50) = 174.75175.50) = 174.75

19961996 175175 174.75 + .10(159 174.75 + .10(159 -- 174.75) = 173.18174.75) = 173.18

19971997 190190 173.18 + .10(175 173.18 + .10(175 -- 173.18) = 173.18) = 173.36173.36

19981998 NANA 173.36173.36 + + .10.10((190190 -- 173.36173.36) = 175.02) = 175.02

Exponential smoothing

84

86

88

90

annual beer consumption ES (a=.1) ES (a=.2) ES (a=.3) ES (a=.4)

17/2/2009 SCM 31

74

76

78

80

82

84

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Trend Projection

annual beer consumption

86

88

90

annual beer consumption

17/2/2009 SCM 32

74

76

78

80

82

84

86

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

r = 1r = 1YY

XX

YYii

= = aa + + bb XXii

^̂

Linear Regression

- New products

- Life cycle products

- Season products

- Trendy products

17/2/2009 SCM 33

r = .89r = .89 r = 0r = 0

XX

YY

XX

YY

XX

YYii

= = aa + + bb XXii

^̂YY

ii= = aa + + bb XX

ii

^̂

- Trendy products

YY a i+ b Xi = + Error

Error

Linear Regression Model

17/2/2009 SCM 34

X

^i i

Error

Observed value

Y a b X= +

Regression line

Prediction based on experience,

74

76

78

80

82

84

86

88

90

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

annual beer consumption

17/2/2009 SCM 35

experience, theory or

assumptions

0,00

2,00

4,00

6,00

8,00

10,00

12,00

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Year average tempature

Exponential smoothing versus ‘temperature’ hypothesis

(1992-2003)

85

86

87

Ex

po

ne

nti

al

sm

oo

thin

g (

a=

.4)

10,50

11,00

11,50

Ye

ar

av

era

ge

te

mp

atu

re

17/2/2009 SCM 36

80

81

82

83

84

78 80 82 84 86 88 90

annual beer consumption

Ex

po

ne

nti

al

sm

oo

thin

g (

a=

.4)

8,00

8,50

9,00

9,50

10,00

78 80 82 84 86 88 90

annual beer consumption

Ye

ar

av

era

ge

te

mp

atu

re

In this case:Moving Averages performs best

0,94

0,96

0,98

1

Correlation with actual trend

17/2/2009 SCM 37

0,84

0,86

0,88

0,9

0,92

MA (n=4) MA (n=3) MA (n=2) MA (n=1) ES (a=.1) ES (a=.2) ES (a=.3) ES (a=.4)

Policy data - facilities

• Production planning

• Capacity planning

• Resource planning

17/2/2009 SCM 38

� Way of working� Formalized procedures� Quality systems� Scenarios� Make or buy

A framework for supply chain decisions – finalCustomers Facilities Suppliers

Products Customer value –portfolio

Product aggregation

Supplier portfolio

17/2/2009 SCM 39

Place Inbound transportation network

Shop floor design Outbound transportation network

Price Margins Direct/indirect costs

External transfer prices

Policy Demand forecasting

Quality management

Make-or-buy

Supply Chain Decision Database support integrative SCM

“(…) In most instances, more than 80% of the data in transactional databases are irrelevant to decision making. Data aggregations and other analyses are needed to transform the remaining 20% or less, into useful information in the supply chain decision database (…)”

17/2/2009 SCM 40

chain decision database (…)”

– What are the overall consequences of reducing 1 or n suppliers?

– What are the overall consequences of increasing 1 or n employees?

– What are the overall consequences of gaining 1 or n customers?