How to run a BI project? Friday, 10 December 2010
Transcription
How to run a BI project? Friday, 10 December 2010
How to run a BI project? Friday, 10 December 2010 But what about success? Friday, 10 December 2010 What is this about? • Look at approaches to projects • Unpinned by management techniques • Aided by best practices Friday, 10 December 2010 What is this about? Methodology Project Management techniques Best practices Friday, 10 December 2010 Also about an unsung hero... ...or how an old guy in a loud shirt can helps us Friday, 10 December 2010 How is this going to work? Slide-uments? Friday, 10 December 2010 How is this going to work? Friday, 10 December 2010 How is this going to work? Friday, 10 December 2010 How is this going to work? Friday, 10 December 2010 How is this going to work? This is about me talking... Friday, 10 December 2010 Who am I? Friday, 10 December 2010 Friday, 10 December 2010 Should you believe me? • Not Prince 2 certified • Not a Scrum Master • This is not a technical project manager piece • I see this in a number of organisations • I see this over a number of years • I did a Software Engineering degree Friday, 10 December 2010 What is a project? • Requirements • Resources • People • Software • Hardware • Time Friday, 10 December 2010 But what about failure? Friday, 10 December 2010 Wikipedia 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Unrealistic or unarticulated project goals Inaccurate estimates of needed resources Badly defined system requirements Poor reporting of the project's status Unmanaged risks Poor communication among customers, developers, and users Use of immature technology Inability to handle the project's complexity Sloppy development practices Poor project management Stakeholder politics Commercial pressures Friday, 10 December 2010 So people have Methodologies... Friday, 10 December 2010 Methodology • Waterfall • Agile • Spiral Model Friday, 10 December 2010 Waterfall (Winston Royce) Friday, 10 December 2010 Friday, 10 December 2010 Agile Friday, 10 December 2010 Agile • Embrace change • Develop software in short amounts of time • Multiple development cycles • Emphasis on face-to-face communication Friday, 10 December 2010 Agile Principles • Communication • Feedback • Simplicity • Courage • Respect Friday, 10 December 2010 In practice • Storyboards • Scrums • People standing up in meetings • Sprints • Backlog Friday, 10 December 2010 Good or bad? • Do people adhere to it too rigidly? • Does it suit Business Intelligence? Friday, 10 December 2010 Spiral Model Friday, 10 December 2010 Friday, 10 December 2010 Friday, 10 December 2010 Friday, 10 December 2010 But before we even start... Friday, 10 December 2010 ...it can go wrong Friday, 10 December 2010 Planning and Estimation • “An estimate is not a contract” • Commercial pressures of estimation • Bottom up or top down? Friday, 10 December 2010 Detailed Estimation Approach • Choose the team • Kickoff meeting • Individual preparation • Estimation session • Assemble tasks • Review results Friday, 10 December 2010 Build a Plan Friday, 10 December 2010 Stages of a project • Requirements Analysis • Data Analysis • Design • ETL build • OBIEE build • Testing • Rework • Release Friday, 10 December 2010 Stages of a project Month 1 Discovery Month 2 Month 3 Analysis Review 1 Iteration 1 Development Review 2 Iteration 2 Review 3 Friday, 10 December 2010 Cycle 1 rework + Development Friday, 10 December 2010 Discovery Requirements Analysis Data Analysis Architecture Discovery High level design Planning Infrastructure Friday, 10 December 2010 Iteration 1 Selection Detailed Design ETL Development OBIEE Development Iteration 1 Unit/Smoke Testing Release SIT/UAT Friday, 10 December 2010 Iteration n Selection Rework Detailed Design ETL Development Iteration n OBIEE Development Unit/Smoke Testing Release SIT/UAT Friday, 10 December 2010 Project Roles • Project Manager • Business Analyst • Data Analyst • Dimensional Modeller • ETL Developer • OBIEE Developer • Trainer • Evangelist Friday, 10 December 2010 Requirements • Elicit requirements • Document requirements • Define acceptance criteria Friday, 10 December 2010 Data Analysis • Evaluate availability of source data • Profile data and assess data quality • Define error handling rules • Prototype ETL using SQL Modelling Friday, 10 December 2010 Design • Define dimensional model • Define ETL flow • Define ETL frequency Friday, 10 December 2010 ETL build • Source and Target database objects • Mapping or Interfaces • Orchestration layer • Security Friday, 10 December 2010 OBIEE build • Physical mappings • Logical/business model • Presentation layer • Reports, dashboards, alerts, scorecards etc • Security Friday, 10 December 2010 Testing • Unit testing • Smoke testing • Integration testing • User acceptance testing • Volume testing Friday, 10 December 2010 Rework • See previous steps.... Friday, 10 December 2010 Release • Automated • Repeatable Friday, 10 December 2010 Iterate... Friday, 10 December 2010 Friday, 10 December 2010 How to ensure success? Friday, 10 December 2010 Management techniques • Risk Management • Configuration Management • Release Management • Providing visibility Friday, 10 December 2010 Management techniques Selection Detailed Design ETL Development OBIEE Development Unit/Smoke Testing Release Iteration 1 SIT/UAT Risk mananagement Configuraiton mananagement Release mananagement Visibillity Friday, 10 December 2010 Risk management • Capture and register (ACRI logs) • Communicate • Ownership • Assign resources Friday, 10 December 2010 Configuration Management Friday, 10 December 2010 Release Management Friday, 10 December 2010 Visibility • Adaptive tools • Visible tasks • Easy documentation • Reporting Friday, 10 December 2010 Best practices • Involving the business • Acceptance criteria • SQL Modelling • Reporting tool abstraction • Testing • Training (post implementation support) Friday, 10 December 2010 Involving the business • Involving the business • Co-location • What is their role? Friday, 10 December 2010 We must be able to identify success Friday, 10 December 2010 Early definition of acceptance criteria Friday, 10 December 2010 = early visibility of data to business Friday, 10 December 2010 Prototyping • Evolutionary or throwaway? • SQL modelling • Reporting tool abstraction Friday, 10 December 2010 SQL modelling Friday, 10 December 2010 Reporting tool abstraction Friday, 10 December 2010 Re-aligning OBIEE physical layer Presentation Layer Common Enterprise Information Layer Prototype Customer Dimension Friday, 10 December 2010 Prototype Sales Fact Prototype Product Dimension Re-aligning OBIEE physical layer Presentation Layer Common Enterprise Information Layer Prototype Actual Customer Dimension Friday, 10 December 2010 Prototype Sales Fact Prototype Product Dimension Re-aligning OBIEE physical layer Presentation Layer Common Enterprise Information Layer Prototype Actual Customer Dimension Friday, 10 December 2010 Prototype Sales Fact Prototype Actual Product Dimension Automated Testing • Test-driven development • Automation and smoke testing • Regression testing • Volume testing • Test driven development Friday, 10 December 2010 Training • Ensuring developers are up to speed • Ensuring end-user and tester are up to speed • Post implementation support Friday, 10 December 2010 What is this about? Methodology Project Management techniques Best practices Friday, 10 December 2010 Conclusion • Look at approaches to projects • Unpinned by management techniques • Aided by best practices Friday, 10 December 2010 Questions? Friday, 10 December 2010