Question: Question: Quality and Productivity Productivity measurement
How do you know how good your facility is?
Littlefield Game
What is the goal?
Maximize Shareholder Value! Or Economic Profit EP = NOPAT – Capital Cost
Increase Net Revenue Increase NOPAT Decrease Operating Expense
Increase Sales Revenue Reduce CGS Selling
SERVICE
Types of measures
n
Distribution OPERATIONS Production Overhead Reduce Working Capital Reduce Investment INVENTORY BUILDINGS, EQUIPMENT, LAND
n n n n n
Economic Profit
Reduce Capital Reduce Capital Cost
Customer satisfaction Asset utilization Operating costs Quality Cycle time Productivity
Reduce Cost of Capital
(Recommended from Integrated Distribution Managementby Gopal and Cypress)
1
What is benchmarking?
n
Importance of benchmarking
n
n
n
A standard by which something can be measured or judged (American Heritage Dictionary ) The continuous process of measuring products, services and practices against the toughest competitors or those companies recognized as industry leaders (David Kearns, CEO of Xerox) The search for industry best practices that lead to superior performance (Robert Camp, author)
n
Nearly 80% of top executives surveyed felt that benchmarking was an imperative to success. 95% of the companies surveyed who felt that benchmarking was an imperative, did not know how to do it (from a survey from Financial World )
Why benchmark?
n n
Three perspectives of benchmarking
n
n n
To compare ourselves to our competitors To compare our practices to the practices of other firms (which may or may not be in the same business) To examine the effect of a change in the current practice (e.g. quality improvement) To identify improvement opportunities To assist in budget allocation decisions
n
n
n
Internal - focuses on the operations of a single company. External - looks outside the firm’s industry (but uses same processes). Competitive - looks at firms conducting business in the same industry.
Productivity
n
Classic definition
n
n
“Productivity reflects results as a function of effort” (Kendrick, productivity guru ) As productivity improves then either greater results are achieved for the same level of effort, or using less effort the same results are maintained.
Productivity is defined as: The ratio of output achieved over the level of effort required to achieve that output.
2
Efficient vs. effective
n
Productivity improvement
n
In the area of productivity, we want to be both efficient and effective.
n n
Efficient - to do things right Effective - to do the right things
Three steps to productivity improvement are:
n n n
n
It is of course possible to do one well and not the other.
Performance Measurement Benchmarking Action Plan
From the Productivity Improvement Guidelines for the Value Chain
Performance measurement
n
Benchmarking
n
A productivity reporting system measures operational performance by various “key measures” of customer service, manpower, equipment and material utilization n Measures must be properly structured to avoid internal goals that conflict with business objectives n Key is appropriate accurate information
From the Productivity Improvement Guidelines for the Value Chain
The search for best practices that lead to world class performance.
n
The search can be within the organization, the industry or any other business Present performance is compared against “best practice” to identify the “gap” in performance This “gap” indicates potential improvement opportunities
n
n
From the Productivity Improvement Guidelines for the Value Chain
Action plan
n
Key measures
n
n
Improvement opportunities are prioritized and action plans developed to improve customer service and reduce costs. “Best practice” is not a finite entity, it is a constantly moving set of parameters which improve with the continuous application of these three steps.
Business objectives:
n
customer service, asset management, operational efficiency, long range planning run/shift, day, period (week, month, year) supervisor, department manager, senior management
n
Reporting intervals:
n
n
Users:
n
From the Productivity Improvement Guidelines for the Value Chain
From the Productivity Improvement Guidelines for the Value Chain
3
Example performance measures
n
Financial measures (production)
Production Asset Turnover = (Revenue Production Expenses)/(Production Asset Value) Production Costs per Order Production Cost Ratio = (Total Production Expenses)/Revenue
People Cost Speed Quality
Staffing Setup Time
Product
Service
Inventory Operations
Safety Stock Turnover Rate Fill Rate Forecast Accuracy
Supplier
n n
Warranty Transaction Design Cycle Lead Time % Correct Orders Order Acquisition
Utilization Transaction Cycle Time % Scrap Process Flexibility Lead Time % Correct Orders Vendor Selection
% Scrap Reliability Features
Innovation Training
Ratio -based benchmarks
n
Important ratio properties
n
n
Typically, managers use benchmarks of simple ratios since they are easy to compute. Manufacturing examples:
n n n n
n n
The ratios should be consistent with corporate objectives The ratios should be unbiased The ratios should be customer focused
pieces/hour errors/million cost/piece WIP/shift
How about the following measures?
n n n n n
Cases produced per shift Cases produced per line Cases produced per labor hour Operating cost per dollar of revenue What about the following measures from the Productivity Improvement Guidelines for the Value Chain?
Net Units Produced % Mechanical = efficiency Design Rate Scheduled hours x 100
Best in Class (BIC) Target: Best in Class (BIC) Target: 80% to 90% is considered ”BIC" 80% to 90% is considered ”BIC"
4
Net Units Produced Line Utilization = Design Rate Paid Hours x 100%
Target BIC is 70% to 80% Target BIC is 70% to 80%
Data collection
n
Who is better?
n
n
The key step in any benchmarking study is the data collection component Several considerations must be made including
n n n
n
amount and accuracy of data required cost to obtain data units of measure
Company 1 produces 125 units of output per day while using 25 units of capital and 25 units of labor Company 2 produces 250 units of output per day while using 50 units of capital and 50 units of labor
Returns to scale
n
Constant returns to scale (CRS)
n
Defines how the output changes when the inputs are all multiplied by a constant
If we multiple each input by the same factor, then the output goes up by that factor. Otherwise we have variable returns to scale (VRS).
E.g. if 100 units of labor (L) and 50 units of capital (K) yield 200 units of output (Q), then we have CRS if X*100 units of L and X*50 units of K yield X*200 units of Q
5
Who is better?
n
Problem
In this case, ratio-based benchmarking doesn’t allow us to compare firm C with the others. Ratio-based benchmarks cannot evaluate the “system” as a whole
Company A B C
Labor 40 50 30
Capital 20 21 40
Output 100 95 100
Weighted measures
n
System-based benchmarking
n
One solution is to apply “weights” based on the importance of each factor: Score = ∑w i xi
In system-based benchmarking, we want to examine the overall (or “system”) operating efficiency and cost efficiency in terms of:
n n n
n
How do we get the w i values?
n
Resources Used Services Delivered Prevailing Prices
The output of system-based benchmarking is an efficiency measure of the firm (or system)
System measure
n
Efficiency
Efficiency is a measure of how well a firm uses it inputs to achieve outputs (the measure is between 0 and 1 were 1 is efficient) The efficiency frontier is defined for a given output, and is the set of firms that use their inputs as efficiently as possible to achieve their outputs (e.g. have a efficiency of 1)
Inputs
• Labor • Capital • Square Ft.
Firm
Outputs
• Units • Profit
n
Efficiency
6
Efficiency frontier
Input 2 Assuming all firms are scaled to the same output
Efficiency score
Input 2
Efficiency =
A A+B
Best In Class
Input 1
B A
Input 1
Virtual firm
n
Two-dimensional example
Input 2
Virtual firm for 2 is made up of 67% of firm 1 and 33% of firm 3.
n
n
The graphical technique does not work in general (only for 2 input, 1 output) A mathematical technique called data envelopment analysis (DEA) will work for any number of inputs and any number of outputs. To determine efficiency of a firm, it constructs a virtual firm based on the other firms in the study in order to compare (using a technique called linear programming)
4 2 1 3
Virtual firm
Input 1
Virtual firm
n
Mathematical Notation
x ij is level of input j for firm i yik is level of output k for firm i θ 0 is the efficiency of reference firm 0 λi is the "intensity"of firm i (i.e. the amount of the firm used to make up virtual firm) n is the number of firms
Imagine a new competitor who studied the industry closely and took from each firm only its very best ideas and most efficient technology
7
Mathematical Formulation (CRS)
min θ 0 subject to :
Mathematical Formulation (VRS)
min θ 0 subject to :
∑λ x
i= 1 n
n
i ij
≤ x ojθ 0 ≥ y ok
∀j ∀j
∑λ x
i=1 n i=1 n i
n
i ij
≤ x ojθ 0 ≥ y ok
∀j ∀j
∑λ y
i i= 1
ik
∑λ y ∑λ
i=1 i
ik
λi ≥ 0 ∀i
(Note: there is a typo in handout!)
= 1, λi ≥ 0 ∀i
(Note: there is a typo in handout!)
Example Formulation
Example Formulation (for firm A)
min θ A
Company A B C
Labor 40 50 30
Capital 20 21 40
Output 100 95 100
subject to: 40λ A + 50λB + 30λ C ≤ 40θ A 20λ A + 21λ B + 40λC ≤ 20θ A 100λ A + 95λB + 100λC ≥ 100 λA , λB , λC ≥ 0
Example Formulation (for firm B)
min θ B subject to: 40λ A + 50λB + 30λC ≤ 50θ B 20λ A + 21λ B + 40λ C ≤ 21 B θ 100λ A + 95λ B + 100λC ≥ 95 λA , λB , λC ≥ 0
n
Proxies
Given the model, we next need to determine how the inputs and outputs are measured In many cases, the desired input or output may not be directly measured. In this case we need to develop a surrogate measure or proxy
n
8
DEA issues
n
Issues continued
n
n
Often times the collected data is subject to measurement error which can greatly effect results In the absence of measurement error, DEA is an upper-bound estimate of “true” efficiency (i.e. if we added a new firm to the study, then the existing firm scores would stay the same or get worse).
n
n
In order to estimate efficiency “reasonably” well you should have enough firms in your study (about 3 times number of inputs time number of outputs) Deciding which inputs and outputs are appropriate is key and much time should be spent here Often times proxies must be used
Why does benchmarking fail?
n n n n n n n n n
The process is not driven by results Unclear objectives Leading team is not credible Lack of stakeholder involvement/buy -in Poor up-front planning Organization becomes defensive Study becomes delaying mechanism Misinterpretation of data Seen as solution - not the starting point
n
Sun Tzu (From The Art of War , 473BC): “If you know your enemy and know yourself, you need not fear the result of 100 battles. If you know yourself but not the enemy, for every victory gained you will also suffer a defeat. If you know neither the enemy nor yourself, you will succumb in every battle.”
9