The Business Benefits of CMMI at NCR Self Service European SEPG Conference 16th June 2004 Walter Scott Walter.Scott@ncr.com
Agenda � Context: NCR Self Service, Scotland � Process Improvement Journey � Benefits gained in Software Process Improvement � Challenges in changing culture � Benefits gained using CMMI � Reasons for success � Issues faced � Questions and Answers NCR Confidential
NCR ~ 30,000 NCR Employees World Wide Financial Retail Teradata Data Warehousing Self-Service Store Automation NCR Confidential
Self Service: Global Presence and Customer Focus Shipments to 32% share of More than 500,000 130 countries Installed base Installed NCR ATMs Manufacturing Global customer Global professional Scotland-Canada-China services services India-Brazil Softw are Hardw are NCR Services NCR Confidential
Self Service in Scotland Main Campus - Dundee ~ 1300 Employees ~ 500 in R&D 40 Project Managers ~ 90 active projects (Duration 2 months to 18 months)
Self Service Process Improvement Journey 2001 2002 2003 2004 2000 1998 1999 CMM CMM CMM CMMI CMMI CMMI CMM CMM CMM CMMI CMMI CMMI Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 rating rating rating rating rating target rating rating rating rating rating target Software Process Process Improvement Improvement using S/W using CMMI CMM (Scope All Development and (Scope Software Development) Product Management) Mini Assessment NCR Confidential
Self Service Process Improvement Journey 2001 2002 2003 2004 2000 1998 1999 CMM CMM CMM CMMI CMMI CMMI CMM CMM CMM CMMI CMMI CMMI Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 rating rating rating rating rating target rating rating rating rating rating target Software Process Process Improvement Improvement using using CMMI Software CMM (Scope All Development and (Scope Software Development) Product Management) Mini Assessment NCR Confidential
Benefits Gained by Software Process Improvement NCR Confidential
CMMI and Predictability Level Predicted Performance Result Characteristic Probability Continuous process Productivity Target N-z 5 Optimizing improvement & Quality Time/$/... Probability Process measured and 4 Target N-y Quantitatively controlled Managed Time/$/... Probability Process characterized 3 Target N-x Defined for the organization and R is proactive Time/$/... I Probability Process characterized Target N+a S 2 Managed for projects and is often reactive K Time/$/... . Probability Target N Process unpredictable, 1 Initial poorly controlled, reactive Time/$/... NCR Confidential
Software Schedule Variance Level 1 to 2 70% CMM Level 1 60% Projects Planned between 1992 to 1996 % of projects . 50% 0% of projects were 40% within +/- 25% 30% 85% of projects were > 75% late 20% Probability 10% Target N 0% -25 to +25 26 to 75 76 to 125 126 to 175 176 to 225 > 225 Time/$/... % overrun in schedule from initial plan 70% 60% CMM Level 2 Projects Planned between 1997 to 1999 % of projects . 50% 40% 60% of projects were within +/- 25% 30% 75% of projects were 20% < 75% late 10% Probability Target N+a 0% -25 to +25 26 to 75 76 to 125 126 to 175 176 to 225 > 225 Time/$/... % overrun in schedule from initial plan . NCR Confidential
Software Schedule Variance CMM Level 2 to 3 ATC Variance - Before & After - Original Baseline Average UCL LCL Delta Days BEFORE AFTER 350.00 Sigma 1.44 Sigma 2.30 300.00 44.00 Average Average 18.47 250.00 200.00 149.19 150.00 100.00 56.15 50.00 0.00 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 -50.00 -19.21 -61.19 -100.00 NCR Confidential
Software Development Cycle Time Software Development Cycle Time 90 80 70 60 Weeks 50 40 30 20 10 0 <1998 1998 1999 2000 2001 Year of Delivery NCR Confidential
Other Benefits � Greater visibility of project status to senior management � Expectation that schedule will be met � Managed process now the foundation for continuous process improvement based on measurement � Improved employee satisfaction � Engineering projects are now measured against realistic targets � Improved understanding of risks associated with projects NCR Confidential
Software CMM to CMMI 2001 2002 2003 2004 2000 1998 1999 CMM CMM CMM CMMI CMMI CMMI CMM CMM CMM CMMI CMMI CMMI Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 Level 1 Level 2 Level 3 Level 2 Level 2 Level 3 rating rating rating rating rating target rating rating rating rating rating target Software Process Process Improvement Improvement using S/W using CMMI CMM (Scope All Development and (Scope Software Development) Product Management) Mini Assessment NCR Confidential
Changing culture ‘As was’ situation 2001 in Engineering Target culture � Reactive � Proactive � Process Averse � Process Discipline � Wide ranges of practices � Consistent practices � Limited lessons learned � Lessons Learned every time � Ad hoc decision making � Structured and fact based decision � ‘Finger in the air’ estimation making � Data driven estimation � Process owners in name only � Active Process Management � Only a few involved in Process � Majority involved in Process Improvement Improvement � Unpredictable processes � Predictable processes NCR Confidential
Challenges of changing culture -1 � Starting point - ‘As was’ situation 2001 ⊗ Compressed time frame “reach Level 3 in 2 years” ⊗ Lack of Project Management discipline in Hardware Development large overruns poor visibility of portfolio status inconstant process usage Issue Management style - reactive Risk Management almost non existent ⊗ Short term ROI required for a medium to long term type investment However we were able to build on…. √ S/W team achieved Software CMM Level 3 in June 2001 √ Most Project Managers had gained Masters Certificate in Project Management √ Strong Engineering Discipline √ Project Management Office started 2001 √ 6 Sigma program underway in NCR NCR Confidential
Challenges of changing culture -2 � Convincing the management team to invest the effort √ The business case for change was built on previous success in using Software CMM locally and in other parts of NCR � Getting peoples time to work on improvement when they are too busy fire fighting √ Actively involved Senior Management √ Project Management Office dual role of helping projects out of crisis and organisational improvement √ Trained Project Managers in CMMI and sought their involvement in planning and action teams √ Teams lead by experienced “ change agents” involving the practitioners at required stages. • to understand the issues • come up with solutions • validate solution / working practices & pilot new practices NCR Confidential
Challenges of changing culture - 3 � Getting people to execute a process for the first time √ Training, facilitation and support √ Using ‘evangelists’ to promote good practice √ Templates and guidance √ Communication and reminders √ Planned Assessments helped to provide additional motivation √ Project Management Certification Programme � Achieving Institutionalisation √ Process Measures and Management Review with senior management √ Setting targets and reporting against them regularly √ Process Audits and assessments NCR Confidential
Challenges of changing culture - 4 � Keeping the momentum going √ Communication - emails, cascaded presentations, community events, easy to access Process, templates and guidelines through web √ Regular Lessons Learned forums √ Rewards for good process improvement suggestions √ Appraisals, focus the attention NCR Confidential
Some Results to date Development Variance Improvement 180 Average Days Variance from Plan 160 140 120 100 80 60 40 20 0 < 2000 2000 H1 2000 H2 2001 H1 2001 H2 2002 H1 2002 H2 2003 H1 2003 H2 Period Project Planned NCR Confidential
Project Cycle Time Improvement Development Project Cycle Time Reduction 120% % of average pre 2000 cycle time 100% 80% 60% 40% 20% 0% < 2000 2000 2001 2001 2002 2002 2003 2003 2000 H1 H2 H1 H2 H1 H2 H1 H2 Module Integration Projects 100% 68% 84% 71% 62% 57% 61% 60% 55% 100% 95% 49% 60% 51% 95% 51% 47% 34% Module/Device Projects Period Project Planned NCR Confidential
Other Benefits √ Broader scope of SEI CMMI vs. Software CMM → From Requirements elicitation to final validation √ Common language, templates and practices across disciplines √ Sufficient consistency process in process execution to enable measurement and process improvement √ Improved rigour in Project Management √ Project Managers now thinking about ways to improve their processes √ Better understanding of impact of changes √ Improved usage of lessons learned √ Formal project governance structure with more consistent management control and insight into status NCR Confidential
Recommend
More recommend
