SUPPLIER SELECTION - My LIUCmy.liuc.it/MatSup/2009/Y90102/3 SUPPLIER SELECTION... · supplier...
Transcript of SUPPLIER SELECTION - My LIUCmy.liuc.it/MatSup/2009/Y90102/3 SUPPLIER SELECTION... · supplier...
Percorso di Eccellenza in PROJECT MANAGEMENT
SUPPLIER SELECTIONSUPPLIER SELECTION
Guido MICHELI & Enrico CAGNO02.2399.405602.2399.4054
Guido [email protected]@PoliMI.it
2
AgendaAgenda
Date Object Tutorial Examples Teamwork
Thursday 29/10Morning (4h) Project Procurement
(focus point)
Thursday 05/11
Morning(4h) Portfolio Approaches +Afternoon (4h)
Tuesday 17/11
Morning(4h) Supplier Selection +Afternoon (4h) & Vendor Rating
(hints)
Wednesday 02/12
Afternoon (4h) Project Work (follow-up)
Guido JL Micheli © LIUC - Supplier Selection
Guido JL Micheli © LIUC - Supplier Selection 3
Team WorkTeam Work
+ ResultsCommunication &Discussion
Guido JL Micheli © LIUC - Supplier Selection 4
Selecting the most proper supplier and assess its performance
Supplier Selection: the goalSupplier Selection: the goal
Getting the required supply from a supplier at an affordable level of risk, using resources efficiently
Guido JL Micheli © 5LIUC - Supplier Selection
A Supplier Selection process can consist of:
• Pre-qualification• Qualification• Selection
• Vendor Rating
Supplier Selection Process
Supplier SelectionSupplier Selection
Guido JL Micheli © 6LIUC - Supplier Selection
LIUC - Supplier Selection 7
Supplier SelectionSupplier Selection
Guido JL Micheli ©
A Supplier Selection (& VR) process must be • Effective• Efficient
USE PORTFOLIO APPROACHES!
Supplier SelectionSupplier Selection
Introduction• Supplier Selection has strongly to do with
DECISION MAKING
Guido JL Micheli © 8LIUC - Supplier Selection
Decision Making is a support for choicesChanging car? Use the Total Cost of Ownership to compare alternatives
Decision Making is intended to justify choicesTry to explain to your classmate why you have chosen a car/bike instead of another
Decision Making has to be consistent and repeatebleYou always need convincing arguments…
Decision modelling (elements) Decision process structure
Supplier SelectionSupplier Selection
Guido JL Micheli © 9LIUC - Supplier Selection
Elements of a Decision
Objective
Strategies (alternative approaches)
States of the world (constraints)
Pay-offs
Supplier SelectionSupplier Selection
Guido JL Micheli © 10LIUC - Supplier Selection
Typically: multi-parameter models (Scoring Models, AHP), used for selection/qualification
Which kind of benefits/limitations?
SS under CertaintySS under Certainty
cost
Suppliers
pa
pb
BA
Guido JL Micheli © 11LIUC - Supplier Selection
Risk is MEASURABLE
Typically: Total Cost of Ownership & advanced methods (REBaSS), used for selection
A range of values as outcome…
SS under RiskSS under Risk
cost
Suppliers
CaMax
BA
CaMin
CbMax
CbMin
Guido JL Micheli © 12LIUC - Supplier Selection
States of the world
Probability can make things different…
How can we make a decision?
SS under RiskSS under Risk
0.6 0.3 0.1
High Medium Low
A 60 80 100
B 80 40 50
supplier
Probability and cost
Guido JL Micheli © 13LIUC - Supplier Selection
Expected Value
EVa = 70EVb = 65
The difference is smoother…
SS under RiskSS under Risk
0.6 0.3 0.1
High Medium Low
A 60 80 100
B 80 40 50
supplier
∑ ⋅= ii xpEV
Probability and cost
Guido JL Micheli © 14LIUC - Supplier Selection
Variance
Va = 180Vb = 345
Let’s consider both EV and V…
SS under RiskSS under Risk
0.6 0.3 0.1
High Medium Low
A 60 80 100
B 80 40 50
supplier
22ii
2 EVxp −⋅=σ ∑
Probability and cost
Guido JL Micheli © 15LIUC - Supplier Selection
‘B’ is no more so PLAIN to be better!
Try to make risk-efficient decisions… by means of decision trees
SS under RiskSS under Risk
Risk
cost
A
B
Guido JL Micheli © 16LIUC - Supplier Selection
DECISION TREESThe problem
Supplier A asks for 80, B for 50.
‘B’ selected supplier?
SS under RiskSS under Risk
Guido JL Micheli © 17LIUC - Supplier Selection
DECISION TREESThe decision involves many scenarios…
Supplier A asks for 80, B for 50.‘A’ is a best in class company, and it is unlikely to go bankrupt. If ‘A’ is
late (Chance 0.1), the cost rises up to 130. No quality problems are expected.
‘B’ is an emergent supplier. If it goes bankrupt (Chance 0.1), the total cost is 250. If not, it is likely to be late 0.4, and the cost is estimated to be 100. In any case, quality problems are expected 30% probable, with re-making extra-costs of 40.
SS under RiskSS under Risk
Guido JL Micheli © 18LIUC - Supplier Selection
DECISION TREESModelisation
KEY
Decision
Chance event (beyond our control)
SS under RiskSS under Risk
Guido JL Micheli © 19LIUC - Supplier Selection
DECISION TREESModelisation
SS under RiskSS under Risk
A
C 0.1 130
C 0.9 80
B C 0.1 250
C 0.9
C 0.4 100 + X
C 0.6 50 + X
C 0.3 40
C 0.7 0
Guido JL Micheli © 20LIUC - Supplier Selection
DECISION TREESConclusion
Expected cost of
‘A’ = 85‘B’ = 98.8
Even if the price of ‘B’ was noticeably lower
SS under RiskSS under Risk
Guido JL Micheli © 21LIUC - Supplier Selection
Uncertainty is NOT measurable, or we do NOT want spend our effort to measure itThe problem
Typically: used for early selection phases (both pre-qual. & qualification)
SS under UncertaintySS under Uncertainty
No idea of probability zones
High Medium Low
A 60 80 100
B 80 40 50
C 90 60 70
D 80 60 50
Probability and cost
supplier
Guido JL Micheli © 22LIUC - Supplier Selection
MiniMin & MiniMaxOpposite attitudes
SS under UncertaintySS under Uncertainty
No idea of probability zones
High Medium Low
A 60 80 100
B 80 40 50
C 90 60 70
D 75 60 50
Probability and cost
supplier
Guido JL Micheli © 23LIUC - Supplier Selection
Coefficient of optimismMiddle attitudes
SS under UncertaintySS under Uncertainty
MiniMin MiniMax Weighted average
A 60 100 76
B 40 80 56
C 60 90 72
D 50 75 60
cost, provided alpha=0.6
supplier
MiniMax)1(MiniMin ⋅α−+⋅α
Guido JL Micheli © 24LIUC - Supplier Selection
Coefficient of optimismChanging attitudes
Notice risk efficient decisions!
SS under UncertaintySS under Uncertainty
MiniMax)1(MiniMin ⋅α−+⋅α
cost
Alpha = 0.6
5040
B
A
60
8075
90100
C
D
Guido JL Micheli © 25LIUC - Supplier Selection
Guido JL Micheli © 26LIUC - Supplier Selection
Scoring Models criteria - ApplicationScoring Models criteria - Application
Scoring Models – Overall TeamworkScoring Models – Overall Teamwork
TeamworkSame groups of 4/5 people• 2 are buyers• 2/3 are competing suppliers (A, B, C)• RULES & ROLES: Group 1 is going to BUY something from Group 7;
Group 1 is going to SELL something to Group 2; …
• 2 buyers describe (a) supply to perform (generic description), (b) supplier selection process, and (c) (weighted) selection criteria consistent with (a)
• 2/3 suppliers (separately) describe their performance (supply and company, i.e. as in vendor rating) in both a quantitative way (1-100 for each criterion, average<=85) and a qualitative one (max 20 lines)
• The group selects the most proper supplier• The whole team comments the results
• Comments• Discussion
Guido JL Micheli © 27LIUC - Supplier Selection
Presentations – Overall TeamworkPresentations – Overall Teamwork
TeamworkSame groups of 4/5 people
• Prepare a 20 mins presentation (assemble of teamworks sessions)
• Contents:1) Portfolio Analysis, Assumptions & Practices 2) General description of items to be purchased (to be revealed to suppliers,
without criteria and weights) 3) Description of Supplier Selection Process selected, Criteria used, Weights
fixed (consistent with PA) (do not reveal that content to the suppliers!) 4) Supplier Choice5) Comments on 1)+2)+3)+4)
Not to be presented: 2/3 Suppliers profiles and offers descriptions (criteria, qualitative description, weights, ...) (to be sent to buyers)
Guido JL Micheli © 28LIUC - Supplier Selection