Chapter 3 Construct Strategy Map & Balanced Scorecard
4.4 System Dynamics Model of the Target Company with Feedback Loop
4.4.1 Strategy Map of the Target Company with Feedback Loop The lines in red and bold are the Feedback
Figure 4.7 Strategy Map of the Target Company with Feedback Loop Source: From this research
Finanacial
Customer
Internal Process
Learning & Growth
Customer Satisfaction
Maintainability
Operation Efficiency
Quality Assurance
Training & Education
Equipement Utilization Readiness
Profit Rate
Employee Capability Realiability
Cost Down
Mission Accomplish Fixed Income
Shareholder Satisfaction
4.4.2 System Dynamics Model of the Target Company w/Causal Feedback Loop
Figure 4.8 Causal Feedback Loop Diagram for the Model of the Target Company Source: From this research
4.4.3 System Dynamics Model of the Target Company with Causal Feedback Loop / Flow Wire
Figure 4.9 Flow Wire Diagram of the Model of the Target Company with Causal Feedback Loop
Source: From this research
Customer
4.4.4 System Dynamics Model of the Target Company with Causal Loop Feedback / Variables Equation
(01) L Customer Satisfaction= INTEG (Satisfaction Var, 0.4) Units: Dmnl
【Customer Satisfaction is a Level Variable that is the integral of Satisfaction Variation with initial value of 0.4.】
(02) R Satisfaction Var=Readiness2Satisfaction* ((TargetCustomerSatisfaction -Customer Satisfaction)/Time2EffectSatisfaction)
Units: Dmnl
【Satisfaction Var is a Rate Variable that adjusts the difference between TargetCustomerSatisfaction and Customer Satisfaction.】
(03) N TargetCustomerSatisfaction=0.85 Units: Dmnl
【Set the Target Customer Satisfaction to be 0.85】
(04) A Time2EffectSatisfaction=3 Units: Month
【The duration of taking the Target Customer Satisfaction into effect】
(05)A Readiness2Satisfaction = WITH LOOKUP (Readiness, ([(0,0)-(10,1)], (0.4,0.4),(0.5,0.45),(0.51,0.46),(0.52,0.47),(0.53,0.48)
,(0.54,0.5),(0.55,0.55),(0.56,0.6),(0.57,0.65),(0.58,0.7),(0.59,0.75),(0.6 ,0.8),(0.61,0.85),(0.62,0.9),(0.63,0.95),(0.64,0.96),(0.65,0.97),(0.66,0.98),(0.
67,0.99),(0.68,1),(0.69,1),(0.7,1),(0.71,1),(0.72,1) )) Units: Dmnl
【A Multiplier for transforming Readiness to Customer Satisfaction】
(06) L Readiness= INTEG (Readiness Var,0.4) Units: Dmnl
【Readiness is a Level Variable that is the integral of Readiness Variation with initial value of 0.4.】
(07)R Readiness Var=Reliability2Readiness *(TargetReainess-Readiness)
/Time2EffectReadiness Units: Dmnl
【Readiness Var is a Rate Variable that adjusts the difference between Target Readiness and Readiness.】
(08) A TargetReadiness=0.64 Units: Dmnl
【Set the Target Readiness to be 0.64】
(09) A Time2EffectReadiness=3 Units: Month
【The duration of taking the Target Readiness into effect】
(10)A Reliability2Readiness = WITH LOOKUP ( Reliability,
([(0.8,0)-(10,0.8)],(0.85,0.4),(0.9,0.45),(1,0.48),(1.1,0.5),(1.2,0.55), (1.3,0.58),(1.4,0.6),(1.5,0.62) ))
Units: Dmnl
【A Multiplier for transforming Reliability to Readiness】
(11)L Reliability= INTEG (Reliability Var, 0) Units: Dmnl
【Reliability is a Level Variable that is the integral of Reliability Var with initial value of 0.】
(12)R Reliability Variability=Efficiency2Reliability*Usage
Rate*((TargetReliability-Reliability)/Time2EffectReliability) Units: Dmnl
【Reliability Variability is a Rate Variable that adjusts the difference between Target Reliability and Reliability.】
(13)A Time2EffectReliability= 5 Units: Dmnl
【The duration of taking the Target Reliability into effect】
(14)A TargetReliability=1.5 Units: Dmnl
【Set the Target Reliability to be 0.64】
(15)A Efficiency2Reliability = WITH LOOKUP (Efficiency,
([(0,0)-(1,2)],(0.3,0.8),(0.4,0.9),(0.5,1),(0.6,1.1),(0.7,1.2),(0.8,1.3) ,(0.9,1.4),(1,1.5) ))
Units: Dmnl
【A Multiplier for transforming Efficiency to Reliability】
(16)A Efficiency=(InitialEfficiency+ActionEffectOnEmployeeEfficiency)-L2E Units: Dmnl
【Efficiency is equal to Initial Efficiency +ActionEffectOnEmployeeEfficiency -(Loading to Efficiency)】
(17)A InitialEfficiency=0.5 Units: Dmnl
【Set the Initial Efficiency to be 0.5】
(18)A ActionEffectOnEmployeeEfficiency=DELAY1I(Strategy Target Action, DelayOnAction , 0 )
Units: Dmnl
【The effect of taking the Action of Improving Employee Efficiency】
(19)A DelayOnAction=2 Units: Month
【The duration of taking the Action of Improving Employee Efficiency into effect】
(20)A Strategy Target Action=STEP( 0.1 , 3 ) Units: Dmnl
【A Step Function that take the Strategy Target Action into effect from the third month and raise to 0.1】
(21) A Efficiency2Maintenance = WITH LOOKUP (Efficiency,
([(0,0)-(2,10)],(0.1,1.9),(0.2,1.8),(0.3,1.7),(0.4,1.6),(0.5,1.5),(0.6,1.4 ),(0.7,1.3),(0.8,1.2),(0.9,1.1),(1,1) ))
Units: Dmnl
【A Multiplier for transforming Efficiency to Maintenance】
(22) A Time2EffectMaintainability= 4 Units: Month
【The duration of taking the Target Maintainability into effect】
(23) L Maintainability= INTEG (MaintainabilityVar,15) Units: Dmnl
【Maintainability is a Level Variable that is the integral of MaintainabilityVar with initial value of 15】
(24)R MaintainabilityVar=
(Efficiency2Maintenance*TargetMaintainability-Maintainability)/Time2EffectM aintainability)
Units: Dmnl
【MaintainabilityVar is a Rate Variable that adjusts the difference between Target Maintainability and Maintainability.】
(25) A TargetMaintainability= 8 Units: Dmnl
【Set the TargetMaintainability to be 8】
(26)A M2L = WITH LOOKUP
(Maintainability,([(0,0)-(20,10)],(7,1),(8,1.1),(9,1.2),(10,1.25),(11,1.3),(12,1.35 ),(13,1.4),(14,1.45),(15,1.5) ))
Units: Dmnl
【A Multiplier for transforming Maintainability to Loading】
(27)A Loading=M2L*Capacity*UsageRate2Loading Units: Hr/Month
【Loading is equal to (Maintainability to Loading) * Capacity * Usage Rate to Loading】
(28)A Capacity="No.OfEmployee"*8*22 Units: Hr/Month
【Set the Capacity to be No.OfEmployee*8(Hours/Day)*22(Days/Month)】
(29)A "No.OfEmployee"=150 Units: People
【Set the No. of Employee to be 150 People】
(30)L OperatingCost= INTEG (CostVar,1e+006) Units: $
【Operating Cost is a Level Variable that is the integral of CostVar with initial value of NT$1000000】
(31)R CostVar=FixedCost+((Loading-Capacity)*70) Units: $/Month
【CostVar is a Rate Variable that is equal to Fixed Cost + (Loading-Capacity)
* 70(Labor Rate: NT$/Hour).】
(32)A FixedCost=1.75e+007 Units: $/Month
【Set the Fixed Cost to be NT$17500000】
(33)L OperatingIncome= INTEG (IncomeVar,0) Units: $
【Operating Income is a Level Variable that is the integral of IncomeVar with initial value of 0】
(34)R IncomeVar=IF THEN ELSE(Readiness>0.54, FixedIncome , FixedIncome-((0.54-Readiness)*FixedIncome))
Units: $/Month
【IncomeVar is a Rate Variable that is a condition: If Readiness>0.54 then get the Fixed Income. But if Readiness<0.54, then the income is equal to Fixed Income-((0.54-Readiness)*Fixed Income)】
(35)N FixedIncome=1.87e+007 Units: $/Month
【Set the Fixed Income to be NT$18700000】
(36)L Operating Profit= INTEG (ProfitVar, 0)
Units: $
【Operating Profit is a Level Variable that is the integral of ProfitVar with initial value of 0】
(37)R ProfitVar=OperatingIncome-OperatingCost Units: $/Month.
【ProfitVar is a Rate Variable that is equal to Operating Income- Operating Cost】
(38)A Usage Rate = WITH LOOKUP (Customer Satisfaction,
([(0,0)-(10,1)],(0.5,0.6),(0.6,0.6),(0.7,0.7),(0.8,0.8),(0.9,0.9),(0.95, 0.95),(0.99,0.96) ))
Units: Dmnl
【The expression of the relationship between Usage Rate and Customer Satisfaction】
(39)A UsageRate2Loading = WITH LOOKUP ( Usage Rate,([(0.4,0)-(10,2)], (0.5,0.8),(0.55,0.85),(0.6,0.9),(0.65,0.95),(0.7,1),(0.75,1.1),(0.8,1.2) ,(0.85, 1.3),(0.9,1.4),(0.95,1.5) ))
Units: Dmnl
【The expression of the relationship between Usage Rate and Loading】
(40)A L2E = WITH LOOKUP (Loading,([(0,0)-(60000,10)],
(10000,0.1),(20000,0.2),(30000,0.3),(40000,0.4),(50000,0.5) )) Units: Dmnl
【The expression of the relationship between Loading and Efficiency】
4.4.5 System Dynamics Model of the Target Company With Causal Feedback Loop / Outcome Plots
1 Dmnl
Customer Satisfaction : Current Dmnl
Efficiency : Current Dmnl
Loading : Current Hr/Month
Maintainability : Current Dmnl
Operating Profit : Current $
Readiness : Current Dmnl
Figure 4.10 Outcome Plot of 6 Indices for the model with causal feedback loop
2 Dmnl
Usage Rate : Current Dmnl
OperatingCost : Current $
FixedIncome : Current $/Month
Figure 4.11 Outcome Plot of 4 Indices for the model with causal feedback loop
4.5 Model Validation and Verification 12 tests are used to validate and verify the model:
(1) Boundary Adequacy Tests
Boundary adequacy tests assess the appropriateness of the model boundary for the purpose at hand. The first step is to determine what the boundary is.
(2) Structure Assessment Tests
Structure assessment tests ask whether the model is consistent with knowledge of the real system relevant to the purpose. Structure assessment focuses on the level of aggregation, the conformance of the model to basic physical realities such as
conservation laws, and the realism of the decision rules for the agents.
(3) Dimensional Consistency
Dimensional consistency is one of the most basic tests and should be among the very first you do. You should always specify the units of measure for each variable as you build your models.
(4) Parameter Assessment
Before deciding how a parameter should be estimated or whether its value is reasonable make sure every constant (and variable) has a clear, real-life meaning.
Next you must decide how to estimate the values of each parameter. The basic choice is formal statistical estimation from numerical data, or judgmental estimation.
(5) Extreme Condition Tests
Models should be robust in extreme conditions. Robustness under extreme conditions means the model should behave in a realistic fashion no matter how extreme the inputs or policies imposed on it may be.
(6) Integration Error Tests
System dynamics models are usually formulated in continuous time and solved by numerical integration. You must select a numerical integration method and time step that yield an approximation of the underlying continuous dynamics accurate enough for your purpose.
(7) Behavior Reproduction Tests
Many tools are available to assess a model’s ability to reproduce the behavior of a system. Most common are descriptive statistics to assess the point-by-point fit.
(8) Behavior Anomaly Tests
Data limitations often mean it is not possible to establish the significance or strength of important relationships or formulations by statistical means. Behavior anomaly tests examine the importance of these structures by asking whether anomalous behavior arises when the relationship is deleted or modified.
(9) Family Member Tests
The family member test asks whether the model can generate the behavior of other instances in the same class as the system the model was built to mimic.
(10) Surprise Behavior Tests
Discrepancies between model behavior and expectation indicate that there are flaws in the formal model, the mental model, or both. Often, of course, discrepancies between model output and your understanding of the system’s dynamics indicate defects in the formal model.
(11) Sensitivity Analysis
Since all models are wrong you must test the robustness of your conclusions to uncertainty in your assumptions. Sensitivity analysis asks whether your conclusions change in ways important to your purpose when assumptions are varied over the plausible range of uncertainty.
(12) System Improvement Tests
The ultimate goal of modeling is to solve a problem. System improvement tests ask whether the modeling process helped change the system for the better. To pass the test, the modeling process must identify policies that lead to improvement, those policies must be implemented, and the performance of the system must actually improve.
4.5.1 Extreme Condition Test Table 4.6 Scenario Extreme Test
Item Scenario Parameters & Descriptions Test 1 Steady State Test No Strategic Improvement Action
Test 2 Overwork Situation Strategic Improvement Action Raise up to 50%
Test 3 Reverse Policy Strategic Improvement Action Down to -10%
Test 4 Prolong Period Simulation Period Extends to 10years 4.5.1.1 Steady State Test
0.8 Dmnl
Customer Satisfaction : Current Dmnl
Maintainability : Current Dmnl
Readiness : Current Dmnl
Reliability : Current Dmnl
Operating Profit : Current $
Loading : Current Hr/Month
Figure 4.12 Outcome Plot of 6 Indices for the model in Steady State test Source: From this research
Outcome Analysis:
From the graph shown above, we can see that Customer Satisfaction, Maintainability, Readiness, Reliability, and Loading are almost remaining the same level. However, the profit can reach to 450 million dollars, and it is fair when the target company do nothing on strategic action. Therefore, the test is in accordance with the normal behavior of regular company performance.
4.5.1.2 Overwork Situation Test
1 Dmnl 0.4 Dmnl 60,000 Hr/Month
16 Dmnl 0.6 Dmnl0 $
0 Dmnl 0 Dmnl 20,000 Hr/Month
14 Dmnl -1 B $
0.4 Dmnl
0 6 12 18 24 30 36 42 48 54 60
Time (Month)
Customer Satisfaction : Current Dmnl
Efficiency : Current Dmnl
Loading : Current Hr/Month
Maintainability : Current Dmnl
Operating Profit : Current $
Readiness : Current Dmnl
Figure 4.13 Outcome Plot of 6 Indices for the model in Overwork Situation test Source: From this research
Outcome Analysis:
From the graph shown above, we can see that Customer Satisfaction and Readiness still can reach their goals while Loading raise up to 60,000 Hr/Month. Efficiency goes up at the beginning and then collapses to 0.1. Maintainability acts just reversely as Efficiency, it goes down at the beginning and then raises up to 15. The profit seriously declines to -1 billion dollars since the cost significantly increases. Therefore, the test proved that if we push too much in efficiency, then it would not reach our goal and act reversely instead. So it matches with the normal behavior of the regular company performance just like we expected.
4.5.1.3 Reverse Policy Test
1 Dmnl 0.4 Dmnl 60,000 Hr/Month
40 Dmnl 200 M $
0.6 Dmnl 0 Dmnl -0.02 Dmnl 20,000 Hr/Month
0 Dmnl -40 M $
0.4 Dmnl
0 6 12 18 24 30 36 42 48 54 60
Time (Month)
Customer Satisfaction : Current Dmnl
Efficiency : Current Dmnl
Loading : Current Hr/Month
Maintainability : Current Dmnl
Operating Profit : Current $
Readiness : Current Dmnl
Figure 4.14 Outcome Plot of 6 Indices for the model in Reverse Policy test Source: From this research
Outcome Analysis:
From the graph shown above, we can see that all of these indicators, except for Efficiency, such as Customer Satisfaction, Maintainability, Readiness, and Loading are acting pretty much the same as basic settings. However, the profit can reach to 150 million dollars, a little lower than 220 million, the original level. Efficiency here goes down to –0.02, the bottom of the coordination. It is reasonable if the target company loose its control of its operation efficiency. Therefore, the test is in accordance with the normal behavior of regular company performance.
4.5.1.4 Prolong Period Test
1 Dmnl 0.4 Dmnl 60,000 Hr/Month
40 Dmnl 1 B $
0.6 Dmnl
0 Dmnl 0.1 Dmnl 20,000 Hr/Month
0 Dmnl -40 M $
0.4 Dmnl
0 12 24 36 48 60 72 84 96 108 120
Time (Month)
Customer Satisfaction : Current Dmnl
Efficiency : Current Dmnl
Loading : Current Hr/Month
Maintainability : Current Dmnl
Operating Profit : Current $
Readiness : Current Dmnl
Figure 4.15 Outcome Plot of 6 Indices for the model in Prolong Period test Source: From this research
Outcome Analysis:
From the graph shown above, we can see that Customer Satisfaction, Maintainability, Readiness, Efficiency, and Loading nothing change in their behaviors till 120 month.
The only difference is the profit can reach to almost 1 billion dollars, and it is rational that the target company would act like this way if it keeps its operation stable from the beginning. Therefore, the test proved that it is in accordance with the normal behavior of regular company performance.
Chapter 5 Conclusions and Suggestions
The purpose of this study is to research the interaction between strategic goals planning and performance measurement indices when enterprises carry out the Balanced Scorecard. Most of the enterprises use the way of bottom up to determine whether their departments or employee could achieve the corporation’s strategic goal.
However, the balanced scorecard use the way of top down to plan the strategic goal and try to find the KPI among those causal relationship. In this way, there are some side effect and unintended consequence happening in the implementing process of strategic goal planning and performance measurement indices constructing. Then, it ends up with “counterintuitive behavior”.
Therefore, it is very important to realize how to choose the proper indices when implementing balanced scorecard. The objective of this study is to help the
enterprises how to take the side effect and unintended consequences into account when they are planning and implementing balanced scorecard. The side effect and unintended consequence are considered as dynamic complex issues that are significantly influence the behavior of the whole system, By adopting the System Dynamic as the tool of resolving and clarifying the dynamic complex issues and applying it on the target company in this study, I hope to contribute to current managerial research and practical application.
5.1 Research Discoveries and Conclusions
This study sorted out a set of planning and designing procedure for the dynamic enterprise's strategic goals and performance measurement as Fig.3.1. The aforesaid way is to launch the balanced scorecard and find the performance measurement indices from the strategy map, Assist enterprises to set up their own strategy map and describe every step of designing performance measurement indices on balanced scorecard, Then constructing the systematic dynamics model in accordance with the way mentioned above and run the simulation to dynamically test every performance measurement indices. By constantly observing the result of the simulation, recheck the original model design and tune the corporate strategic plan, and then adjust the performance measurement indices, help the growth of enterprises.
According to aforesaid documents and literatures, apply the target company into the model set up by following the system dynamics modeling process. And implement the testing, analysis, and discussing. This study sum up the discoveries and conclusions are as follows:
(1) There exists the side effect and unintended sequences in the causality of the strategic goal and actions when implementing the balanced scorecard.
(2) The simulation outcome of the system dynamics model in executing the strategic goal action from balanced scorecard may take it as a reference to plan and adjust the strategic goals for enterprises.
(3) By analyzing the strategic goals causality and various kinds of possible side effects of the balanced scorecard, the results of this study may help enterprises to find out and conclude the key performance indices.
(4) Utilizing the test and simulation of the system dynamics, find out the leading index in the system
(5) According to the research of this study, set up the basic system dynamics model for balanced scorecard.
5.2 Contribution to Managerial Practice
(1) Sort out and construct the procedure and practice of the system dynamics model simulation from depicting strategy map and launching the balanced scorecard.
(2) The side effects of the negative feedback loop caused by the measurement items and performance indices among the four perspectives of balanced scorecard.
Since the methods and tools of the existing balanced scorecard did not consider the side effect, systematic dynamics happen to be the perfect tool to carry out the designing, simulating, and testing of the policy.
(3) Remind the top management to pay more attention on diagnosing and
understanding the management of the performance indices when they implement the strategic actions at any time. Also find out the potential key performance index and analyze its influence to what extent, and put forward the solution as early as possible.
(4) Practical suggestions for the target company based on the result of this study:
The simulation result of the target company's model in this study shows that the employee's work efficiency improves, then drive the rising of the reliability and declining of the maintainability. The improvement of readiness rate cause the customer satisfaction be relatively higher, also increase the utilization rate of airplanes. Increase the utilization rate of airplanes, the scheduled maintenance hours come earlier than regular flight using way, and cause the fatigue of spare parts of airplanes and raise the failure rate. It end up with the increasing in total man-hour of maintenance, then the work loading of the employees is overweight.
The overtime expenses rise obviously as well and staff satisfaction declines too, Since the income of the target company in this study is collected in fix firm price, the main strategic goal for the target company is to minimize the operating cost and maximize the profit rate. In this way, it is unable to reach this goal if staff's satisfaction drops and the turn over rate rises. Nevertheless, the side effect of improving employee's work efficiency reversely causes the employee's work efficiency even worse than the original situation eventually. The target company should be aware that if maintainability and employee's work loading have not been taken care appropriately, then even tremendously improve employee's work efficiency will not make any better for the whole company operation. Therefore, the maintainability and employee's work loading of the target company become their leading indices, must put more efforts on them.
5.3 Research Constraints and Follow-up proposition 5.3.1 Research Constraints
(1) Since it is quite extensive and complicated that the organization promotes the balanced scorecard, it will be too broad if the target company try to include all its units. Therefore, I narrowed down the research scope to maintenance control and management that are the main issues with complicated dynamics behavior.
Through confining the target company organization involved, this study could focus on the specific incident that may be simply apply to different department or different enterprises as a basis to discuss and review. It is worthy of follow-up study.
(2) Because of the difficulty to collect the information needed in this research, and the restrictions of research time, conduct an analysis through the systematic thinking, and then collect the further materials, such as the mental model of the top
management, generic strategic goal, balanced scorecard, and the secondary industry information. Moreover, implement the part of operation procedure of the target company and make a few assumptions to construct system dynamics model. The objective is to inspect the side effect and the unintended
consequences of the model. Through this modeling and simulation in computer, we can perform the unit consistency measuring, extreme value test, and policy improvement test. However, not all of the quantified data in this study have been tested and calibrated because of the restrictions mentioned above.
5.3.2 Follow-up proposition
Through analyzing the target company in this study, we can realize the procedure of constructing the balanced scorecard and strategic goals. For those enterprises that
Through analyzing the target company in this study, we can realize the procedure of constructing the balanced scorecard and strategic goals. For those enterprises that