ST.

THOMAS’ COLLEGE OF ENGINEERING AND TECHNOLOGY

Software Engineering Code: OEC-AIML 801C

COST BENEFIT ANALYSIS

A. K. SIROMONI
STCET
 Lecture on COST BENEFIT ANALYSIS

At the End of the Lecture Students will be able to -

Lecture Objective 1 : Understand Cost Estimation with respect to

Benefits in a Project or Decision

Lecture Objective 2 : Classify the Project for Cost Estimation

Lecture Objective 3 : Estimate the Project according to Classification

A.K.SIROMONI (AIML) 2
 PEOPLE -> PROCESS -> PRODUCT

• A PROJECT SUCCESS DEPENDS ON THE DECISIONS TAKEN ON

PREPARATION OF THE PRODUCT
• PEOPLE DOMAIN DECISIONS ARE DEADLINES, TEAM
SIZE, EFFECTIVE COMMUNICATION, WORK BREAKDOWN,
MAN HOUR ESTIMATES etc
• PROCESS or TECHNICAL DOMAIN DECISIONS ARE METHODS,
TOOLING, DATA, DESIGN, VERSION CONTROL, DOMAIN
KNOWLEDGE etc

Depending on the DECISIONs after Communication

with Client and Internal Discussion , the Cost of the
Project is determined

A. K. Siromoni ( STCET) 3
 PROJECT COST BENEFIT ANALYSIS
FOLLWING POINTS ARE CONSIDERED DURING COST BENEFIT ANALYSIS
COST
• DIRECT COST : SALARIES, SOFTWARE TOOLS, HARDWARE INFRASTRUCTURE,
EXTERNAL SERVICES etc
• INDIRECT COST : INSURANCE, FACILITY RENTALS, UTILITY COSTS etc
• HIDDEN COST : COSTS RELATED TO CUSTOMER SATISFACTION, EMPLOYEE MORALE,
OR OVERALL PRODUCTIVITY
• OPPORTUNITY COST : COSTS RELATED TO CHANGE OF DECISIONS DURING
PREPARATION OF SOFTWARE PRODUCT
• WASTED COST : COSTS RELATED TO WASTED TIME, MISSED POINTS, OVERTIME ETC.
BENEFIT
• DIRECT BENEFIT : THE MEASURABLE BENEFITS FROM A PROJECT, SERVICE, OR
SOFTWARE, SUCH AS REVENUE, PROFIT, OR SALES, ARE KNOWN AS DIRECT BENEFITS.
• INDIRECT BENEFIT : INDIRECT BENEFITS ARE THE QUALITATIVE BENEFITS THAT
CANNOT BE MEASURED BUT PERCEIVE. FOR INSTANCE, CHANGE OF GOVERNMENT
POLICY , DECREASE IN PRICE OF THE RESOURCES, INCREASED BRAND AWARENESS OR
CONSUMER BASE.

A. K. Siromoni ( STCET) 4
 PROJECT COST ESTIMATION
PROJECT COST DEPENDS ON THE SIZE OF THE SOFTWARE AND
CALCULATED FROM THE EFFORT OF THE MAN POWER REQUIERED
FOR IMPLEMENTATION OF THAT SOFTWARE

COST ESTIMATION MODEL

• FUNCTIONAL POINT ANALYSIS
• HEURISTIC TECHNIC ( COST CONSTRUCTION
MODEL ) BASED ON MACHINE LEARNING
• MODELS
• SOFTWARE LIFE CYCLE MODEL ( SLIM )

A. K. Siromoni ( STCET) 5
 • FUNCTIONAL POINT
ANALYSIS

Functional Points (FPs) of an application are the number

of jobs ( Functions ) and their types used in the
applications. Depending on the complexity of the type of
each job , weightage is considered against each function.
Functional Point Analysis shows :
• Detailed User Requirement
• Profit From The Project
• Project Size
• Man Hour Required
• Time Required for Completion
• Complexity of The Project
A. K. Siromoni ( STCET) 6
 • FUNCTIONAL POINT
ANALYSIS
COMPONENTS TO BE CONSIDERED IN FP CALCULATION
1. Transactional Functional Type
External Input (EI): EI processes data or control
information that comes from outside the application’s
boundary. The EI is an elementary process.
External Output (EO): EO is an elementary process that
generates data or control information sent outside the
application’s boundary.
External Inquiries (EQ): EQ is an elementary process made up of
an input-output combination that results in data retrieval.
2. Data Functional Type
Internal Logical File (ILF): A user-identifiable group of logically
related data or control information maintained within the
boundary of the application.
External Interface File (EIF): A group of users recognizable
logically related data allusion to the software but maintained
within the boundary of another software.
A. K. Siromoni ( STCET) 7
 • FUNCTIONAL POINT
ANALYSIS
CALCULATION of
FPs
FP = Count_total * [0.65 + 0.01 * ∑(fj)]
Where Complexity Adjustment Factor (CAF) = [0.65 + 0.01 * ∑(fj)]
Count_total = Total Number of Functions considering weightage
fj = Consideration of 14 General System Characteristics ( GSC )
Value of fj varies from 0 to 70 considering value of j varies from 0
to 14

GSC : ( Data Communication , Distributed Data Processing , Performance,

Heavily used Configuration , Transaction Rate, On-Line Data Entry,
End User Efficiency, On Line Update, Complex Processing,
Reusability, Installation Ease, Operational Ease, Facilitate Change,
Multiple Size )

A. K. Siromoni ( STCET) 8
 • FUNCTIONAL POINT
ANALYSIS
WEIGHTING FACTORS of
FPs
MEASUREMENT
SIMPLE AVERAGE COMPLEX
PARAMETER
NUMBER OF USER
INPUTS 3 4 6
NUMBER OF USER
OUTPUTS 4 5 7
NUMBER OF USER
INQUIRES 3 4 6
NUMBER OF INTERNAL
LOGIC FILES 7 10 13
NUMBER OF USER
EXTERNAL INTERFACES 5 7 10

A. K. Siromoni ( STCET) 9
 • FUNCTIONAL POINT
ANALYSIS
Example CALCULATION of FPs

PROBLEM :
Consider a software project with the following information domain
characteristic for the calculation of function point metric.
Number of external inputs (I) = 30
Number of external output (O) = 60
Number of external inquiries (E) = 23
Number of files (F) = 08
Number of external interfaces (N) = 02
It is given that the complexity weighting factors for I, O, E, F, and N are 4, 5, 4,
10, and 7, respectively. It is also given that, out of fourteen value adjustment
factors that influence the development effort, four factors are not applicable,
each of the other four factors has value 3, and each of the remaining factors
has value 4. The computed value of the function point metric is _____.
[GATE CS 2015]
A. K. Siromoni ( STCET) 10
 • FUNCTIONAL POINT
ANALYSIS
CALCULATION OF TOTAL
COUNT
MEASUREMENT NUMBER OF
WEIGHTAGE TOTAL
PARAMETER FUNCTIONS
NUMBER OF USER
INPUTS 3O 4 120
NUMBER OF USER
OUTPUTS 60 5 300
NUMBER OF USER
INQUIRES 23 4 92
NUMBER OF INTERNAL
LOGIC FILES 08 10 80
NUMBER OF USER
EXTERNAL INTERFACES 02 7 14
TOTAL COUNT 606
A. K. Siromoni ( STCET) 11
 • FUNCTIONAL POINT
ANALYSIS
CALCULATION OF GENERAL SYSTEM
CHARACTERISTICS ( fj )
FACTORS WITH VALUE NUMBER TOTAL

3 4 12
4 6 24
TOTAL ( ∑(fj) ) 36

FUNCTIONAL POINT VALUE

FP = Count_total * [0.65 + 0.01 * ∑(fj)]
= 606 * [0.65 + 0.01 * 36]
= 606 * [0.65 + 0.36]
= 606 * 1.01
= 612.06
A. K. Siromoni ( STCET) 12
 HEURISTIC TECHNIC ( COST CONSTRUCTION
MODEL )
Heuristic means discover. The heuristic technique solves
the problems using approximation for quick and
effective solution.
The most common HEURISTIC TECHNIC for COST BENEFIT
ANALYSYS is COST CONSTRUCTION MODEL ( COCOMO ) .

COCOMO MODEL was suggested by BARRY BOEHM in

1981 and is based on the study of 63 projects

COCOMO MODEL determines estimated effort in person-

month for creating software project in terms of kilo lines
of code ( KLOC )

A. K. Siromoni ( STCET) 13
 COST CONSTRUCTION MODEL ( COCOMO )
COCOMO EFFORT CALCULATION is based on PROJECT TYPE:
i) ORGANIC ii) SEMI DETACHED iii) EMBEDDED
TYPE ORGANIC SEMI DETACHED EMBEDDED
SIZE Project Size Project Size is Project Size is big.
relatively small medium ( more than 300 KLOC
( 2 to 50 KLOC ) ( 50 to 300 KLOC )
)
TEAM Team comprises Team comprises Team comprises of
of highly of experienced experienced and
experienced and inexperienced
software inexperienced software engineers
engineers with a software with some expert
knowledge of engineers both. personnel .
preparing similar
projects
COMPLEXITY Low Medium High
ENVIRONMEN Less Constraints Moderate Strict Requirements
T Constraints
EXAMPLE Inventory Finance and Air Control System,
A. K. Siromoni ( STCET) 14
Management, Accounting, Instrument Automation
 COST CONSTRUCTION MODEL ( COCOMO )
COCOMO ESTIMATION CAN BE DIVIDED IN THREE TYPES OF
MODELS:
i) BASIC ii) INTERMEDIATE iii) DETAILED
BASIC COCOMO MODEL is simple and quick and used for small and
medium projects . It estimates Effort from estimated KLOC in terms of
Person Month
E = a*(KLOC)b Person Month where E is EFFORT
Tdev = c*(E)d where Tdev is Develop Time in Month
Tp = E / Tdev where Tp is total persons required
a, b, c, d are constants determined by the category of software project given in
below table
Project
a b c d
Category

Organic 2.4 1.05 2.5 0.38

Semi-
3.0 1.12 2.5 0.35
Detached

Embedded 3.6 1.20 2.5 0.32

A. K. Siromoni ( STCET) 15
 COST CONSTRUCTION MODEL ( COCOMO )
Example of BASIC COCOMO MODEL EFFORT CALCULATION

Suppose that a Basic project was estimated to be 40 KLOC (kilo lines of

code). Calculate effort and time
40 KLOC can be considered as ORGANIC Category
So effort = 2.4 × (40)1.05 = ( 2.4 x 48.1 ) ≈ 115 person-month.
dev. time = 2.5 × (115)0.38 = (2.5 x 6 ) ≈ 15 months.
Suppose that a Basic project was estimated to be 250 KLOC (kilo lines of
code). Calculate effort and time
250 KLOC can be considered as SEMI DETACHED Category
So effort = 3.0 × (250)1.12 ≈ ( 3.0 x 485 ) ≈ 1455 person-month.
dev. time = 2.5 × (1455)0.35 = (2.5 x 12.8 ) ≈ 32 months.
Suppose that a Basic project was estimated to be 500 KLOC (kilo lines of
code). Calculate effort and time
500 KLOC can be considered as EMBEDDED Category
So effort = 3.6 × (500)1.20 ≈ ( 3.6 x 1739 ) ≈ 6238 person-month.
dev. time = 2.5 × (6238)0.32 = (2.5 x 12.8 ) ≈ 41 months.
A. K. Siromoni ( STCET) 16
 COST CONSTRUCTION MODEL ( COCOMO )
INTERMEDIATE COCOMO MODEL
BASIC COCOMO MODEL CONSIDERS NUMBER OF LINES IN CODE
(KLOC) ONLY WITH SOME CONSTANTS DEPENDING ON THE PROJECT
TYPE.

INTERMIDIATE COCOMO MODEL CONSIDERS FIFTEEN OTHER

FACTORS RELATED TO SYSTEM WITH NUMBER OF LINES OF CODE
(KLOC) IN COST ESTIMATION . THESE 15 FACTORS ARE CALLED COST
CLASSIFICATION OF FIFTEEN COST DRIVERS OF INTERMIDEATE COCOMO MODEL
DRIVERS .
Product Attributes Personnel Attributes
1. Required software reliability extent 8. Analyst capability
2. Size of the application database 9. Software engineering capability
3. The complexity of the product 10. Applications experience
11. Virtual machine experience
Hardware attributes 12. Programming language experience
4. Run-time performance constraints
5. Memory constraints Project Attributes
6. The volatility of the virtual machine 13. Use of software tools
environment 14. Application of software engineering methods
7. Required turnabout time 15. Required development schedule
A. K. Siromoni ( STCET) 17
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL

E = a*(KLOC)b * EAF Person Month where E is the Effort

Tdev = c*(E)d Month where Tdev is the development Time
Tp = E / Tdev Persons where Tp is total persons required
a, b, c, d are constants determined by the category of software project
given in below

Project
a b c d
Category

Organic 3.2 1.05 2.5 0.38

Semi-
3.0 1.12 2.5 0.35
Detached

Embedde
2.8 1.20 2.5 0.32
d
A. K. Siromoni ( STCET) 18
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL METRICES

The Effort Adjustment Factor (EAF) is determined by

multiplying the effort multipliers associated with each of the
15 factors ( Cost Drivers ) related to the system.
Each factor has weightage value considering the factor
impact in project to be
i) Very low ii ) Low iii) Nominal iv) High v) Very High
vi) Extra High

A. K. Siromoni ( STCET) 19
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL METRICES ( COST DRIVERS )

Product Very Extra

Low Nominal High Very High
Attributes Low High
Required
software
0.75 0.88 1.00 1.15 1.40
reliability
extent

Size of the
application 0.94 1.00 1.08 1.16
database

The complexity
0.70 0.85 1.00 1.15 1.30 1.65
of the product

A. K. Siromoni ( STCET) 20
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL METRICES ( COST DRIVERS )

Hardware Very Very

Low Nominal High Extra High
Attributes Low High

Run-time
performance 1.00 1.11 1.30 1.66
constraints

Memory
1.00 1.06 1.21 1.56
constraints

The volatility
of the virtual
0.87 1.00 1.15 1.30
machine
environment

Required
0.87 1.00 1.07 1.15
turnabout time

A. K. Siromoni ( STCET) 21
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL METRICES ( COST DRIVERS )

Very Very Extra

Personnel Attributes Low Nominal High
Low High High

Analyst capability 1.46 1.19 1.00 0.86 0.71

Software engineering
1.29 1.13 1.00 0.91 0.82
capability

Applications
1.42 1.17 1.00 0.86 0.70
experience

Virtual machine
1.21 1.10 1.00 0.90
experience

Programming
1.14 1.07 1.00 0.95
language experience

A. K. Siromoni ( STCET) 22
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL METRICES ( COST DRIVERS )

Extra
Project Attributes Very Low Low Nominal High Very High
High

Use of software
1.24 1.10 1.00 0.91 0.82
tools

Application of
software
1.24 1.10 1.00 0.91 0.83
engineering
methods

Required
development 1.23 1.08 1.00 1.04 1.10
schedule

A. K. Siromoni ( STCET) 23
 COST CONSTRUCTION MODEL ( COCOMO )

EXAMPLE of INTERMEDIATE COCOMO MODEL

CONSIDER A PROJECT where a Software has to be prepared for

a Company for Inventory Control with Cost Analysis of Each
Material in Inventory

After Consultation with Client , the estimated Lines of Codes are

i) DATABASE INPUT : 20KLOC
ii) Interface Screens : 30KLOC
iii) Program Files : 70KLOC
iv) Testing Files : 10 KLOC
v) User Manual : 40HLOC

Total Job Size is 170 KLOC . So it can be classified as semi detached

type.
The job requires front-end and back-end database development. So
some expertise and experience of some team members are required for
development of Software. So for Cost Estimation , Intermediate
COCOMO Model may be considered.
A. K. Siromoni ( STCET) 24
 COST CONSTRUCTION MODEL ( COCOMO )
EXAMPLE of INTERMEDIATE COCOMO MODEL
Effort Adjustment Factor (EAF)
Product Attributes
Size of the application database Nominal 1
Complexity of the Product Nominal 1
Hardware attributes
Runtime Performance Constraint High 1.11
Personnel attributes
Analyst capability Nominal 1
Software engineering capability High 1
Applications experience High 0.91
Project attributes
Use of software tools High 0.91
Application of software engineering methods Very High 0.83
Required development schedule High 1.10

The Effort Adjustment Factor (EAF ) = 1 X 1 X 1.11 X 1 X 1 X 0.91 X 0.91 X 0.83 X 1.10 = 0.84
A. K. Siromoni ( STCET) 25
 COST CONSTRUCTION MODEL ( COCOMO )

INTERMEDIATE COCOMO MODEL

E = a*(KLOC)b * EAF Person Month where E is the Effort

Tdev = c*(E)d Month where Tdev is the development Time
Tp = E / Tdev Persons where Tp is total persons required

File Size = 170 KLOC EAF = 0.84

a = 3.0 b = 1.12 c = 2.5 d = 0.35

So Effort (E) = 3*(170)1.12 * 0.84 PM

= 3 * 314.85 * 0.84 PM = 793.42 Person Month
Development Time ( Tdev ) = 2.5 * (793.42)0.35 M
= 2.5 * 10.35 M = 26 Month
Total Persons Required( Tp ) = 793.42 / 26 Persons = 31 Persons

A. K. Siromoni ( STCET) 26
 COST CONSTRUCTION MODEL ( COCOMO )

DETAILED COCOMO MODEL

E = a * (Size)^b * EAF * Mode Factors Person Month where E is the Effort

Detailed COCOMO is used for very big projects divided into several
modules. Cost Drivers for each modules are calculated along with mode
factors consisting wider range of parameters like team experience,
development practices, and software complexity etc. At the end COCOMO
is applied to each module and final cost is determined by summation .
The Six phases of detailed COCOMO are:
Planning and requirements
System structure
Complete structure
Module code and test
Integration and test
Cost Constructive model
A. K. Siromoni ( STCET) 27
 COST CONSTRUCTION MODEL ( COCOMO )
Differences between BASIC , INTERMEDIATE and DETALED COCOMO
Factor BASIC INTERMEDIATE DETAILED
Effort = a * (Size)^b * Effort = a * (Size)^b * EAF *
Equation Effort = a * (Size)^b
EAF Mode Factors

Expanded set of more cost

Cost Drivers Not considered Added 14 Cost Drivers
drivers according to module

Cost Consider only KLOC Consider KLOC with 14 Consider KLOC with Mode
Consideration project specific factors Factors and project specific
( EAF ) factors per module

More comprehensive,
Project Basic software Wide range of project-specific
considers development
Characteristics development only characteristics
environment

Complexity is less as Complexity is more as Complexity is maximum as

Basic COCOMO offers Intermediate COCOMO Detailed COCOMO considers
Levels of
a straightforward considers project Mode Factors of Each module
complexity
estimate based on specific factors with along with project specific
size. size. factors and size.

A. K. Siromoni ( STCET) 28