Context
One of the basic challenges many organizations face relates to benchmarking their
performance within the organization as well as outside itself. Internal benchmarking
implies comparison between the performance of different units / activities with
similar units / activities within the organization having similar goals and deploying
qualitatively similar resources. External benchmarking implies comparison with units
/ activities managed by other organizations. This comparison may become challenging
when a multitude of goals need to be achieved using a number of resources.
Data Envelopment Analysis (DEA) was introduced by operations research practitioners
in late 1970s as a means of facilitating such benchmarking. It was later developed
(and deployed) by professionals from varied streams and industries. DEA has opened
up avenues for analysis of cases which have been resistant to other analysis techniques
due to the complex nature of the relationships between inputs and outputs.
What is DEA
Data Envelopment Analysis (DEA) is a data analysis technique for evaluating the
performance of a set of peer entities called Decision Making Units (DMUs). DEA allows
a lot of flexibility in the definition of DMUs. These units may be branches of a
bank, outlets of a food chain, production units of a manufacturing organization,
departments within the same firm, etc.
The core of DEA comprises identification of efficient mix of DMUs towards realizing
the desired goals with the minimal deployment of resources. The resources deployed
by the efficient mix are then compared with the actual resources deployed by a DMU
for realization of specified goals. This comparison highlights whether the DMU under
evaluation is efficient or not.
An Illustration
The deployment of DEA for efficiency analysis and benchmarking may be better understood
with the following illustration.
Consider a set of 4 production facilities of Elecomp Pvt Ltd, an electronics circuit
supplier. The facilities are located at Manesar, Chakan, Jaipur, and Dharuhera.
The components produced by Elecomp are CE100, HL305, TL207, and SK113. All the components
are manually assembled, using contract workforce, from basic materials like cables
and circuit boards using hand-held devices. Elecomp management considers these materials,
along with the manpower at each plant, as the key input elements. Exhibit 1 provides
the total resource consumption for the different plants in March 2009.
|
|
Chakan
|
Dharuhera
|
Jaipur
|
Manesar
|
|
Cables (‘00 m)
|
348
|
123
|
129
|
154
|
|
Circuit Boards (‘000)
|
104
|
106
|
64
|
104
|
|
FTE personnel (no.)
|
276
|
285
|
162
|
210
|
Exhibit 1: Resource consumption at different plants
The number of components produced by the different units is provided in Exhibit
2
|
|
Chakan
|
Dharuhera
|
Jaipur
|
Manesar
|
|
CE100
|
175
|
253
|
148
|
160
|
|
HL305
|
23
|
41
|
27
|
84
|
|
TL207
|
37
|
48
|
35
|
33
|
|
SK113
|
46
|
43
|
27
|
56
|
Exhibit 2: Component-wise production by different manufacturing plants
Based on the above information, DEA may be deployed to analyze which of the units
are operating relatively inefficiently and what should the optimal inputs (keeping
output constant) or outputs (keeping inputs constant) be for the inefficient plants.
For example, a DEA analysis of the Chakan plant in this case shows that Chakan could
have produced its current output using only 90% of the resources consumed currently.
Alternatively, using the same amount of resources that it consumed, Chakan plant
could have produced 2 more CE100 and 37 more HL305 components.
Advantages
Within 30 years of its introduction by Charnes, Cooper and Rhodes, DEA has gained
wide acceptability across the operations research and analytics fraternity. Both
public and private institutions, across industries, have jumped to adopt DEA for
internal and external benchmarking and thus identifying areas of inefficiency as
well as potential for improvement. Such a quick and widespread acceptance of the
approach is due to a unique set of advantages offered by DEA as listed below
- 1. DEA is an empirical (actual data based) approach and hence grounded in actually
achieved / achievable results
- 2. It does not need priori assumptions regarding relationship between inputs and
outputs. Thus there is no need to explicitly specify a mathematical form for analysis
- 3. It can handle multiple inputs and outputs simultaneously and comprehensively
- 4. It can be applied to time-series data as well as single time-horizon data
- 5. The extent and sources of inefficiency may be identified for all units under
analysis
- 6. It has been proven to be useful in uncovering relationships that remain hidden
from other methodologies
It must be noted, in this context, that DEA calculates relative efficiencies of
the plants. Thus, it does not lead organizations to the absolute efficient frontier
but only to the most efficient frontier already achieved by the different DMUs within
the organization. This very fact, in a sense, also highlights the core benefit of
using DEA i.e. the efficiencies outlined by DEA are more practically achievable
than the efficiencies prescribed / estimated by most of the theoretical analysis
models.
Areas of Application
DEA is being deployed across a wide range of industries and functional areas. Some
of the industries where DEA is being applied currently are highlighted in Exhibit
3
Exhibit 3: Deployment of DEA across various Industries
It can be seen that DEA application cuts across industries. Researchers are continuously
bringing in more industries under the ambit of DEA with innovative ways of leveraging
its power for providing relevant results in different industries.
Similarly, the application of DEA cuts across functional areas. Some of the illustrative
applications of DEA in different functional areas are outlined in Exhibit 4.
Exhibit 4: Applications of DEA across various functional areas
Benefits of leveraging DEA
Implementing DEA delivers a host of benefits to organizations
- 1. Identify inefficient units / functions and improvement areas
- 2. Improve resource efficiency of manufacturing and supply chain operations
- 3. Control risks while enhancing financial returns across projects and investment
options
- 4. Identify right sourcing options for RM procurement
- 5. Develop efficient Sales and Marketing strategies and plans
- 6. Continuously improve performance of various functions through Window Analysis
How to Deploy DEA
To deploy DEA in-house, organizations may train managers, and equip them with the
required software / hardware to conduct the analysis. Alternatively, organizations
may approach data analytics firms that specialize in such techniques and have access
to the desired tools.
At Beacon, we assist organizations leverage DEA for continuous improvement and elimination
of inefficiencies. Our data analytics skills, coupled with high end analytics software
and processors, ensure that the power of DEA may be unleashed to improve the performance
of your organization. To know more about how DEA may help your organization, please
contact us.