Scribd
Upload
57 views24 pages

Parametric Estimation For ERP Implementations

Uploaded by

javatwojava
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
57 views24 pages

Parametric Estimation For ERP Implementations

Uploaded by

javatwojava
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 24

Parametric Estimation for ERP

Implementations
Donald M. Beckett
Quantitative Software Management, Inc.
2000 Corporate Ridge, Suite 900
Mclean, VA 22102
Tel: 703 790-0055, Fax 703 749-3795
Email: [email protected]
Web: www.qsm.com

11/21/2011 © Quantitative Software Management, Inc.


Objectives

• Provide conference attendees with a practical


method for estimating the project size of ERP
implementations that is both easy to learn and
apply
• Compare the behavior of ERP implementations to
other business IT projects
 Size vs. Schedule
 Size vs. Effort

(#2)
Outline

• Key differentiators between ERP implementations


and software development
• Sizing ERP implementations
 Determining size
 RICEF objects
 Configuration items
 Normalizing to a common metric
• Estimating ERP implementations

(#3)
Quotations

“Perfection is the enemy of the possible”


- Voltaire (paraphrased)

“Precision is not accuracy”


- William Horton

(#4)
Key Differentiators

• Software projects create code


 Develop new systems
 Modify existing systems
 Are measured (sized) by the functionality they deliver
and/or the code they create
• Software projects may
 Develop interfaces
 Have hardware, network, telecom components
 Convert data
 Have system setup and configuration

(#5)
Key Differentiators

• ERP Implementations have


 Significant system setup & configuration
 Hardware, network, & telecom components
• ERP Implementations may
 Develop interfaces
 Convert data
 Create additional functionality
 Modify existing functionality

(#6)
Determining Size

• Software project size is not how much it costs


nor how long it takes
• Size measures the functionality a software
project delivers
• Parametric estimation (SLIM, COCOMO, etc.)
uses size as a key input to determine cost and
schedule
 Lines of code, function points, requirements, use cases
are traditional size measures
• What size measures capture the functionality of
an ERP implementation?

(#7) 11/21/2011
Sizing ERP Implementations

• ERP Implementation size: two components


 Configurations
 Customizations
• Configurations include parameters, properties,
rules, values, table setup
• Customizations are principally code
• Proportions vary between projects
• ERP sizing must consider both

(#8)
Configurations

• Estimate the number of configuration items (by


category & complexity)
 Best case, worst case, most likely scenarios
• Alternatively, identify number of high level
business processes that must be configured
 SAP Solution Composer is an example
• Normalize them to a common elementary unit
(using gearing factors)

(#9)
Configuration Example: Tables

• Average table has


 3 indices to define
 20 columns to define
 20 data types (one per column)
• Average table (in this example) requires 43
elementary activities (or implementation units) to
create
 Gearing factor of 43

(#10)
Customizations

• RICEF objects: Reports, Interfaces, Conversions,


Enhancements, Forms
• Estimate counts of each item (by complexity)
• Normalize them to a common elementary unit
(using gearing factors)
• Add to normalized configuration items count for
an estimated project size

(#11)
Sample Gearing Factor Table: RICEF
Objects

Component Gearing Factor Number Size


Simple Reports 100 10 1000
Average Reports 200 5 1000
Complex Reports 300 20 6000
Simple Interfaces 320 2 640
Average Interfaces 620 12 7440
Complex Interfaces 1520 1 1520
Simple Conversion 100 2 200
Average Conversions 200 5 1000
Complex Conversions 300 2 600
Simple Enhancements 100 2 200
Average Enhancements 500 1 500
Complex Enhancements 1000 3 3000
Simple Forms 100 2 200
Average Forms 200 15 3000
Complex Forms 300 3 900
Total 27,200

(#12)
But, Does it Work?

• Step 1: Size completed ERP implementations


using configuration items and RICEF objects
• Step 2: Compare trends for Effort, Schedule,
Staffing, and Productivity to trends for Business
IT projects (non-ERP)

(#13)
Schedule

+1σ

Average

-1σ

(#14)
Schedule

Blue lines are


trends for ERP
implementations
Black lines for
Business IT
projects

Overall, smaller ERP


implementations enjoy a slight
schedule advantage. Larger
ones lack this

(#15)
Effort

(#16)
Effort

Small ERP implementations have


a cost/effort advantage while
larger ones are almost identical to
traditional development

(#17)
Average Staff

(#18)
Average Staff

ERP implementations
use slightly smaller
teams for most
projects although the
trends and the
amount of variability
are very similar

(#19)
Productivity Parameter

(#20)
Productivity Parameter

ERP Implementations
are more productive
than Business IT for
smaller projects but
lose their advantage
as size grows

(#21)
Conclusions

• ERP Implementations have very similar behavior


to other Business IT projects
 Schedule, effort, staffing, productivity
• Parametric estimation techniques used for
Business IT projects are applicable to ERP
implementations
• ERP Implementation size can be effectively
estimated using Configuration Items and RICEF
Objects
 Widely used by U.S. government for estimation and
tracking

(#22)
Conclusions

• Although smaller ERP implementation projects


are slightly more productive than traditional
Business IT, the cost of the package should be
included in cost estimates if it is being purchased
• While larger ERP implementations do not enjoy
cost or schedule advantages, larger traditional
Business IT projects have a higher probability of
failure, which must also be considered when
choosing an alternative

(#23) 11/21/2011
Questions?

(#24) 11/21/2011

You might also like