Service Operations

Management

Lecture – 7

Prof. Pinakhi Suvadarshini

Quantitative Methods and
Operations Management
Email: [email protected]
 PROCESS
IMPROVEMENT
Process improvement in service operations management focuses on enhancing the efficiency and
effectiveness of service delivery. This involves:
• Identifying Bottlenecks: Finding and addressing areas where the service process is slowed down or
interrupted.
• Streamlining Processes: Simplifying and optimizing service procedures to reduce waste and
improve speed.
• Implementing Best Practices: Adopting industry best practices to improve service delivery.

Continuous improvement is a way of thinking that needs to be incorporated into a firm’s culture.
• PDCA cycle
• Total Quality Management
• Benchmarking
• Six Sigma
• Process Mapping
 DATA ENVELOPMENT ANALYSIS
Data Envelopment Analysis (DEA) is a performance measurement tool used to evaluate the relative
(DEA)
efficiency of multiple decision-making units (DMUs), by comparing the amount of inputs used to the
amount of outputs produced.

Also called Frontier Analysis because DEA is used to create an efficiency frontier against which the
performance of decision-making units (DMUs) is assessed.

Efficiency Frontier

It is a curve that is a "best practice" line/curve. It shows the highest level of performance achieved by
the best-performing units in a group. If a unit is on this line, it means it’s as efficient as possible given
its resources. A, B, C are on the line: Efficient units

D, F, E : Inefficient units

The efficiency frontier visually

demonstrates the trade-offs and best
possible performance combinations.
 DATA ENVELOPMENT ANALYSIS
Example: One input and one (DEA)
output

Consider a number of bank branches. For each branch we have a single output measure (number of
personal transactions completed) and a single input measure (number of staffs).
Personal Number of Personal transactions
Branch Transactions Staffs per staff member
How then can we compare these branches
and measure their performance using this ('000s) ('000s)
Delhi 125 18 6.94
data?

Kolkat 2.75
a 44 16
Efficiency Ratio = Personal transactions
4.71
per staff member

Mumb 2.09
Delhi has the highest efficiency.
ai 80 17
Branch Relative Efficiency
Chenn
ai 23 11
As Delhi has the highest ratio of 6.94 we can compare all (6.94/6.94) *100 =
Delhi 100%
other branches to it and calculate their relative
efficiency with respect to Delhi. Kolkata (2.75/6.94) *100 = 40%
Mumbai (4.71/6.94) * 100= 68%
The other branches are relatively less efficient at using
their given input resource (staff members) to produce (2.09/6.94) * 100 =
Chennai 30%
output (number of personal transactions) compared to
 DATA ENVELOPMENT ANALYSIS
Example: One input and two (DEA)
outputs

We have a two outputs measure (number of personal and business transactions completed) and a
single input measure (number of staffs).
Personal Business Number of
How now can we compare these Branch Transactions Transactions Staffs
branches and measure their ('000s) ('000s)
performance using this data?
Delhi 125 50 18
Delhi has the highest ratio of Kolkata 44 20 16
personal transactions per staff Mumbai 80 55 17
member, whereas Mumbai has
Chennai 23 12 11
the highest ratio of business
transactions per staff member. Personal transactions Personal business per
Branch per staff member ('000s) staff member ('000s)

One problem with comparison via Delhi 6.94 2.78

ratios is that different ratios can Kolkata 2.75 1.25
give a different picture and it is
difficult to combine the entire set Mumbai 4.71 3.24
of ratios into a single numeric Chennai 2.09 1.09
judgement.
 DATA ENVELOPMENT ANALYSIS
Graphical analysis (DEA)
One way around the problem of interpreting different ratios, at least for problems involving just two
outputs and a single input, is a simple graphical analysis. Suppose we plot the two ratios for each
branch as shown
Delh
The positions on the graph represented by Delhi i
and Mumbai demonstrate a level of performance
which is superior to all other branches. This line is
called the efficient frontier (sometimes also Mumb
referred to as the efficiency frontier). ai

The efficient frontier, derived from the examples Kolkata

of best practice contained in the data we have
considered, represents a standard of performance Chenn
ai
that the branches not on the efficient frontier
could try to achieve.
 DATA ENVELOPMENT ANALYSIS
Quantifying efficiency scores(DEA)
for inefficient DMU's

We say that any branches on the efficient frontier are 100% efficient (have an efficiency of 100%).
Kolkata and Chennai are less than 100%. But how much? Can we assign an appropriate numerical
value?
Consider Chennai, we have: Delh
• number of staff 11 i
• personal transactions ('000s) 23
• personal transactions per staff member (23/11)
= 2.09 Mumb
• business transactions ('000s) 12 ai
• business transactions per staff member (12/11)
= 1.09 Kolkata

For Chennai the ratio personal transactions : Chenn

ai
business transactions = (23/12) = 1.92, i.e. there
are 1.92 personal transactions for every business
transaction.

By definition this figure of 1.92 is also the ratio of:

A line is drawn
personal from the
transactions perorigin through Chennai’s
staff member : business current position extending outwards. This line
shows all possible
transactions positions
per staff member Chennai could move to if it adjusted its staff levels while keeping the same
ratio of personal to business transactions (1.92)
 DATA ENVELOPMENT ANALYSIS
Quantifying efficiency scores for inefficient DMU's
(DEA)
Hypothetical Improvement:
For example were Chennai to operate the same level of
output with 9 staff we would have:
•number of staff 9 Delh
•personal transactions ('000s) 23 i
•personal transactions per staff member (23/9) = 2.56 BES
•business transactions ('000s) 12 T
•business transactions per staff member (12/9) = 1.33 Mumb
ai
•personal transactions per staff member : business
transactions per staff member = (2.56/1.33) = 1.92
Kolkata

You can hopefully see from the diagram above that the Chenn
point corresponding to personal transactions per staff ai
member = 2.56 and business transactions per staff
member = 1.33 lies on the line from the origin through
the current position of Chennai.

The point "Best" on the diagram indicates where Chennai

would be on the efficiency frontier if it managed this
staffing and transaction balance effectively, achieving
 DATA ENVELOPMENT ANALYSIS
Quantifying efficiency scores for inefficient DMU's
(DEA)
Then in DEA we numerically measure the (relative)
efficiency of Chennai by the ratio:
(Length of line from origin to Chennai /Length of line
from origin through Chennai to efficient frontier) * 100 Delh
i
Similarly it can be done for Kolkata. BES
T
Mumb
ai
There are a number of ways by which Chennai can move
towards that point. It can:
• reduce its input (number of staff) whilst keeping its Kolkata

output constant (an input target); or Chenn

• increase both its outputs, retaining the current ai
personal : business (business mix) ratio of 1.92 whilst
keeping its input (number of staff) constant (an output
target); or
• do some combination of the above.

The other way is to use an optimization technique as

shown in next slide.
 DATA ENVELOPMENT ANALYSIS
Multiple Inputs and Multiple(DEA)
Outputs

Inputs Outputs
IT
Operatin Personal Business Customer
Number of Systems
Branch Staff g Costs Investmen Transactio Transaction Satisfactio
('000s) ns ('000s) s ('000s) n Score
t ('000s)
Delhi 18 120 75 125 50 8.5
Kolkata 16 80 50 44 20 7.0
Mumbai 17 100 65 80 35 8.0
Chennai 11 60 40 23 12 6.5
Evaluating any number of DMU's, with any number of inputs and outputs:
• requires the inputs and outputs for each DMU to be specified;
• defines efficiency for each DMU as a weighted sum of outputs [total output] divided by a weighted sum of
inputs [total input]

Why Weights?
• Different branch might prioritize different inputs and outputs based on their operational strategies and
local demands.
E.g. Delhi might place a higher emphasis on IT Systems Investment (Input 3) due to its focus on digital
banking services.
Kolkata might focus on minimizing operating costs (Input 2) due to its smaller scale
 DATA ENVELOPMENT ANALYSIS
Let , with , be the efficiency ratio of (DEA)
DMU k, where is the total number of DMU’s being evaluated.

Let , with , be a coefficient (weights) for output , where is the total number of output types considered. The variable is a
measure of the relative increase in efficiency with each unit increase of output value.

Let , with , be a coefficient (weights) for input , where is the total number of input types considered. The variable is a measure
of the relative decrease in efficiency with each unit increase of input value

Let be the number of observed units of output generated by service unit during one time period.

Let be the number of actual units of input used by service unit during one time period.

Objective: To find out the weights so that the efficiency is maximized

for each DMU.

Non-linear Programming

Subject to, In solver GRG Non-Linear

∀
,
 DATA ENVELOPMENT ANALYSIS
One way to convert it into Linear Programming
(DEA)

Subject to,

∀
,
 DATA ENVELOPMENT ANALYSIS
An innovative drive-in-only burger
(DEA) chain has established six units in several different cities. Each unit
is located in a strip shopping center parking lot. Only a standard meal consisting of a burger, fries,
and a drink is available. Management has decided to use DEA to improve productivity by identifying
which units are using their resources most efficiently and then sharing their experience and
knowledge with the less efficient locations. Table 1 summarizes data for two inputs: labor-hours and
material dollars consumed during a typical lunch hour period to generate an output of 100 meals sold.
Normally, output will vary among the service units, but in this example, we have made the outputs
equal to allow for a graphical presentation of the units’ productivity.
Service Material
Unit Meals Sold Labor-Hours Dollars
1 100 2 200
2 100 4 150
3 100 4 100
4 100 6 100
5 100 8 80
6 100 10 50
 DATA ENVELOPMENT ANALYSIS
(DEA)
Labor- Material
Service Unit Meals Sold Hours Dollars
1 100 2 200
2 100 4 150
3 100 4 100
4 100 6 100
5 100 8 80
6 100 10 50

Service units S1 , S3 , and S6 have been joined to form an efficient-production frontier of alternative
methods of using labor hours and material resources to generate 100 meals.

Points on the efficiency frontier should not be dominated by any other points.

S2 is dominated by S3 because S3 is using less Material dollars, hence S2 is not there in the frontier
line.
 DATA ENVELOPMENT ANALYSIS
(DEA)
Subject to
Subject to Subject to

Weights are
non-
negative in
each case.

Subject to Subject to Subject to

 DATA ENVELOPMENT ANALYSIS
(DEA)Efficiency
Service Rating (E) Relative Labor- Relative Material
S1, S3 and S6 are Reference Set Hour Value
Unit (Relative Value (V(2))
efficient and to be efficiency) (V(1))
taken as benchmark.
S1 1.000 NA 0.1667 0.0033
S1 (0.2857)
S2 0.857 S3 (0.7143) 0.1428 0.0028

S3 1.000 NA 0.0625 0.0075

S3 (0.7778)
S4 0.889 S6 (0.2222) 0.0555 0.0067

S3 (0.4545)
S5 0.901 S6 (0.5454) 0.0568 0.0068

S6 1.000 NA 0.0625 0.0075

For unit S4 , each unit decrease in labor-hours results in an efficiency increase of 0.0555. For unit S4
to become efficient, it must increase its efficiency rating by 0.111 points. This can be accomplished by
reducing labor used by 2 hours (i.e., 2 hours × 0.0555 = 0.111).

Note that with this reduction in labor-hours, unit S4 becomes identical to efficient unit S3.
 DATA ENVELOPMENT ANALYSIS
S1, S3 and S6 are (DEA)
efficient and to be Relative Labor- Relative Material
Service Efficiency
taken as benchmark. Unit Rating (E) Reference Set Hour Value Value (V(2))
(V(1))
Column 3 tells which S1 1.000 NA 0.1667 0.0033
unit should
considered as S1 (0.2857)
S2 0.857 S3 (0.7143) 0.1428 0.0028
benchmark.
S3 1.000 NA 0.0625 0.0075
For S4, the S3 (0.7778)
benchmarks are S3 S4 0.889 S6 (0.2222) 0.0555 0.0067
and S6. The values in
bracket represents S3 (0.4545)
S5 0.901 S6 (0.5454) 0.0568 0.0068
the relative weight
assigned to that S6 1.000 NA 0.0625 0.0075
efficient unit in
This is obtained from the shadow price of S4.
calculating the
efficiency rating for
S3 and S6 contributes 77.78% and 22.22% respectively to the benchmarking process of S4.
S4.
To improve its efficiency, S4 should focus on strategies and practices employed by both S3 and S6​but
give more emphasis to S6​(since it has a higher weight).
 DATA ENVELOPMENT ANALYSIS
For S4, I want to create a (DEA) Relative Relative
Service Efficiency Reference
composite mix of both the Labor-Hour Material Value
Unit Rating (E) Set
Value (V(1)) (V(2))
inputs.
S1 1.000 NA 0.1667 0.0033
Meals: 100 S1 (0.2857)
S2 0.857 0.1428 0.0028
S3 (0.7143)
Labor Hours:
S3 1.000 NA 0.0625 0.0075
0.7778*4+0.2222*10 = 5.3
S3 (0.7778)
S4 0.889 0.0555 0.0067
Material: S6 (0.2222)
0.7778*100+0.2222*50 = 88.9 S3 (0.4545)
S5 0.901 0.0568 0.0068
S6 (0.5454)
For S4, initially it was Labour S6 1.000 NA 0.0625 0.0075
hours: 6 and Material: 100.

So, S4 has to reduce 0.7 units

of labor hours and 11.1 units of
Material to reach the frontier
line.