High Maturity Implementation in ERP
Engagements
                -Avinash Bharj
                 Mitta Rout
               ...
High Maturity Implementation in ERP Engagements
1st International Colloquium on High Maturity Best Practices
Bangalore, 21...
ERP Engagements – Candidate for HM Implementation


ERP engagements form a major part of vendor organisation’s business & ...
ERP Life Cycle (SAP) @ Capgemini

 Project Preparation
                                                                   ...
Sizing – A Challenge

 No industry functional sizing method (IFPUG FP, UCP) available for SAP ABAP development

 Size is...
Sizing: ABAP - SAP Unit


                                                  X


                                          ...
HM Practices for SAP ABAP Development – Case Study


 SEI CMMI v1.2 (Dev) HM practices, as per L4 / L5, were designed &  ...
Identification of Improvement Goals (for engagement)

                Vision – BO – QPPO                                  ...
Identification of Critical Sub-processes for Statistical Mgmt.
                   PPM (multiple linear regression
        ...
Process Composition
  To select most optimum option for executing a sub-process in order to meet QPPO SL at max certainty...
PPM Usage – During Planning
What-If Analysis          Initial Prediction




What-If Analysis

  To identify the target v...
PPM Usage – During Planning………..
Initial Prediction

  Values obtained from “What-If Analysis” were fed in PPMs (main PPM...
PPM Usage (object Level) – During Execution

 Predict Y based on known X(s) at start of the object (done for each object)...
Statistical Sub-process Management

 Understand critical sub-process variations and possible causes
 For “special” cause...
CAR (Process, Product) & OID

                  Org. level “Delivery Improvement Team” (DIT) – delivery & QA resources
  ...
Improvement Results & Business Impact
Improvement Results (engagement level)

QPPO (Y)          UOM        Y09 Improvement...
Thanks




avinash.bharj@capgemini.com
mitta.rout@capgemini.com

                                                         ...
Click here for more presentations on



        CMMI High Maturity
          Best Practices
            HMBP 2010
        ...
Upcoming SlideShare
Loading in …5
×

CMMI High Maturity Best Practices HMBP 2010: High Maturity Implementation in ERP Engagements by Avinash Bharj and Mitta Rout

2,538 views

Published on

High Maturity Implementation in ERP Engagements by Avinash Bharj and Mitta Rout (Capgemini India Pvt. Ltd.) presented at
1st International Colloquium on CMMI High Maturity Best Practices held on May 21, 2010, organized by QA

Published in: Technology, Business

CMMI High Maturity Best Practices HMBP 2010: High Maturity Implementation in ERP Engagements by Avinash Bharj and Mitta Rout

  1. 1. High Maturity Implementation in ERP Engagements -Avinash Bharj Mitta Rout (Capgemini India Pvt. Ltd.)
  2. 2. High Maturity Implementation in ERP Engagements 1st International Colloquium on High Maturity Best Practices Bangalore, 21st May 2010 Avinash Bharj (SQA Head) & Mitta Rout (Vice President – QA) Capgemini India Private Limited © 2010 Capgemini– All rights reserved
  3. 3. ERP Engagements – Candidate for HM Implementation ERP engagements form a major part of vendor organisation’s business & revenue Solutions to be competitive w.r.t cost & schedule without compromising on quality  Sustain existing accounts  Higher proposal to win ratio (add new accounts) User needs require roll-out of customised versions built on top of vanilla version  Need to demonstrate improvements release-over-release (development process optimisation) Productivity & Quality improvements expected by Senior Management in line with Business Objectives (business sustenance & expansion) Improve Productivity Improve Quality Cost Saving High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 3
  4. 4. ERP Life Cycle (SAP) @ Capgemini Project Preparation ABAP Object BUILD Lifecycle Business Blueprint Phase Tech Spec. Realisation Phase Major part of the Lifecycle ABAP Coding Unit Testing Final Preparation Phase ABAP Code Rework Go Live & Support Phase Form Object Unified Project Management Report Solution Manager FRICE Objects Interface Conversion Enhancement  ERP customisation (ABAP Dev.) can be considered as a small CSD engagement  BUILD activities (ABAP Dev.) are repeated for each object  BUILD activities are mostly done at offshore  Significant gains can be obtained overall by optimising BUILD related sub-processes High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 4
  5. 5. Sizing – A Challenge  No industry functional sizing method (IFPUG FP, UCP) available for SAP ABAP development  Size is critical in computing Productivity & Quality metrics, and driving improvements  Size is also critical from an estimation perspective  As part of Quality Journey, Capgemini designed a “proprietary” method in year 2007  SAP ABAP Object size (across FRICE) is estimated as “ABAP - SAP Unit” (SAP-U)  Is being used for computing:  BUILD Productivity (expressed as Efforts per SAP-U)  Defect Density (expressed as Defects per SAP-U)  Availability of SAP-U, and past data on Productivity & Quality metrics, provided Capgemini a strong foundation for applying HM practices High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 5
  6. 6. Sizing: ABAP - SAP Unit X X X X X X Input Parameters SAP Units Complex High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 6
  7. 7. HM Practices for SAP ABAP Development – Case Study  SEI CMMI v1.2 (Dev) HM practices, as per L4 / L5, were designed & Improvement Goals implemented for SAP ABAP development for FRICE objects > 40 Hrs Identification estimated efforts (for BUILD) Critical Sub-process  For such objects, it was possible to perform “initial predictions” and Identification apply “in-process controls” at sub-process level (during execution)  Availability of data points was not a concern, as SAP customisation Process Composition engagements involve huge number of FRICE objects  Hence, understanding current performance & variations for sub- PPM Usage – During processes, applying improvements & measuring results within same Planning & Execution engagement was possible Statistical Sub- process Management CAR (Process, Product), OID Improvement Results & Business Impact High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 7
  8. 8. Identification of Improvement Goals (for engagement) Vision – BO – QPPO Objectives (Quality, mapping Cost, Schedule) Organisation Client Organisation QPPO improvement goals were identified for Y09 & Y10 Vision Business QPPO (Y) UOM Current Y09 Improvement USL Objective (BO) Performance (eng. Target (over past data) current perfrm.) Increase Expand existing Reduce Defects Mean & Std. Dev. Mean – 11% As per Revenue (Top accounts by Delivered DD per SAP was calculated Std. Dev. – 50% Org. Line) building client (post UT) Unit confidence Improve Increase Improve Person Mean & Std. Dev. Mean – 50% As per Gross profitability Productivity Days was calculated Std. Dev. – 50% Org. Operating (BUILD) per SAP Profit (GOP) Unit Engagement’s past data was used to finalise the QPPO(s) improvement goals for the engagement ANOVA, Dunnet’s, Pareto techniques used Product CAR Goals  Defect Type  10% mean reduction on “Functional” & “Standard” DD Proactive steps to reduce defect injection  50% reduction on Std. Dev.  Defect Cause Code  10% reduction on current distribution for “Improper Coding / Implementation” & “Inadequate Design” High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 8
  9. 9. Identification of Critical Sub-processes for Statistical Mgmt. PPM (multiple linear regression technique); Y = f(X1, X2, X3) Organisation PPB (for the Xs & Ys) QPPO PPM Name PPM Applicable Y Sub-process Sub-process Measures Level Segment (sample list only) (X) Improve BUILD Main SAP FRICE BUILD Effort  Code Creation  Coding Effort per SAP Productivity Productivity Objects per SAP Unit  Requirements Unit (interim Y) Review  Requirement Ambiguity Index Coding Interim SAP FRICE Coding Effort  Technical Spec.  Designer Skill Index Productivity Objects per SAP Unit Creation  Design Reviewer Skill  Technical Spec. Index Review Critical sub-processes were identified for the engagement using Sensitivity Analysis Some of the sub-process selection criteria: X  Significant Impact on Y (Sensitivity Analysis)  “Controllable” by engagement team  Data collection (X) possible; no MS-Error “Crystal Ball” used for Simulation / Sensitivity Analysis High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 9
  10. 10. Process Composition  To select most optimum option for executing a sub-process in order to meet QPPO SL at max certainty level  “Quality” objective was given higher priority than “Productivity” objective – considering client needs  Crystal Ball was fed with org. & engagement data (FRICE), and engagement’s QPPO SL to generate options  Rank # 3 option was found to be most suitable  Some of the sub-process options selected for FRICE objects:  Tech. Spec. creation post requirements intake check using “intake checklist”  Peer as-well-as SME review for Tech. Spec.  Peer as-well-as SME review for ABAP code  Unit testing also to include Functional test cases Requirements Decision Variables 75% Probability or more < 1.17 70% Probability or more < 3.40 Rank Solution # Delivered Defect Density Build Productivity Code creation _FT code_creation_wo_FUT 1 39 96.39% 89.54% 0.00 1.00 2 60 96.39% 89.44% 0.00 1.00 3 47 96.39% 89.39% 0.00 1.00 4 19 96.39% 84.28% 1.00 0.00 High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 10
  11. 11. PPM Usage – During Planning What-If Analysis Initial Prediction What-If Analysis  To identify the target value for each sub-process measures (X) in order to achieve QPPO (Y) target value  These X(s) were a part of the PPM associated with QPPO(s)  “What-If Analysis” was performed using “Crystal Ball” during overall engagement planning / re-planning  Best possible solution was obtained by running 10000 trails of simulation  Mean value for X(s) was obtained from engagement’s past data & same used to select the best solution  “Controllability” of sub-processes by engagement team was also critical factor in the selection High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 11
  12. 12. PPM Usage – During Planning……….. Initial Prediction  Values obtained from “What-If Analysis” were fed in PPMs (main PPM  Interim PPM)  Check was done to see if “Prediction Intervals” (PI) were within SL of the QPPO  Trade-off done for X(s) overlapping between “Productivity” & “Quality” PPM(s), and risks were identified  Action plans were put in place to achieve the selected target value of X(s) (i.e., KT from designer to tester) High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 12
  13. 13. PPM Usage (object Level) – During Execution  Predict Y based on known X(s) at start of the object (done for each object)  If PI > USL, perform “What-If-Analysis” and determine & apply “mid-course corrections”  Compare “actual Y” with “predicted Y” and determine reason for variations High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 13
  14. 14. Statistical Sub-process Management  Understand critical sub-process variations and possible causes  For “special” cause of variation, action taken (to achieve sub-process stability)  I-MR Control Chart for time series data, Box Plots for non time series data (e.g., Skill Index) High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 14
  15. 15. CAR (Process, Product) & OID  Org. level “Delivery Improvement Team” (DIT) – delivery & QA resources Organisation  Past SAP ABAP Dev. engagements defect data analysed; frequently occurring:  Defect Type Product CAR  Defect Cause Code  Root cause determined to reduce injection of defects  SAP Solution Manager – “Code Inspector” made mandatory  OID initiative originated from CAR  Improvement goals set for Y09 & Y10 for “Defect Type” & “ Defect Cause Code”  Engagement’s product CAR goals identified – “Defect Type” & “Defect Cause Code” Engagement  Proactive steps taken to reduce defect injection  “Code Inspector” CAR (Process,  “Binary Logistic” PPM for predicting probability of defect occurrence w.r.t Skill Product)  CAR done for both process & product defects  Process CAR examples;  Sub-process Capability Deficiency  EV & SV outside SL(s)  Post improvement results validated thru’ Hypothesis tests, Capability analysis High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 15
  16. 16. Improvement Results & Business Impact Improvement Results (engagement level) QPPO (Y) UOM Y09 Improvement Y09 Actual Hypothesis Target (over past Improvement (over Test perfrm.) past perfrm.) Passed? Reduce Defects Mean – 11% Mean – 90% Yes Delivered DD per SAP Std. Dev. – 50% Std. Dev. – 71% (post UT) Unit Improve Person Mean – 50% Mean – 55% Yes Productivity Days Std. Dev. – 50% Std. Dev. – 79% (BUILD) per SAP Unit  Significant reduction in identified “defect-type” & “defect cause-code” were achieved; these were also statistically validated Usage of SAP Solution Manager “Code Inspector” helped significantly reduce “Standard” defect-type  Increase in client’s confidence in Capgemini’s “quantitative” approach towards managing & executing large engagements Business Impact  Client awarded Global Support contract (application support for all countries) High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 16
  17. 17. Thanks avinash.bharj@capgemini.com mitta.rout@capgemini.com High Maturity Implementation In ERP Engagements 1st International Colloquium on HMBP (Bangalore, May 21st 2010) © 2010 Capgemini– All rights reserved Version 1.0 / Slide # 17
  18. 18. Click here for more presentations on CMMI High Maturity Best Practices HMBP 2010 organized by QAI Click here

×