Making Sense Of Your Role On An Agile Project

Transcription

1 Making Sense Of Your Role On An Agile Project Source : Evan Dennis - Unsplash.com

2 About Equinox IT Equinox IT is New Zealand s leading independent IT consultancy, delivering software development, consulting and training services. Established in 1995, we inject fresh thinking to solve tough business problems. Equinox IT has over 60 staff based in Wellington and Auckland.

3 A familiar story (in a galaxy not so far away) Our systems are a mess after 8 prior major IT projects We will build a new Deathstar and learn the lessons of the past Our systems are a mess after 9 prior major IT projects We will build a new Deathstar and learn the lessons of the past

4 Project success remains elusive 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Succeeded Challenged Failed Courtesy of the Standish Group

5 Key principle risk scales non-linearly with size 120% 100% 23% 11% 25% 4% 11% 80% 42% 38% 45% 60% 62% 59% 40% 68% 20% 0% 55% 58% 44% 27% 18% 7% 3% Large Agile Large Waterfall Medium Agile Medium Small Agile Small Waterfall $10m+ Waterfall $1m or less Successful Challenged Failed Courtesy of the Standish Group, Chaos Manifesto 2015

6 Agile is an approach to minimising risk Programming defect found through Pair Programming Design or programming defect found through Test Driven Development (TDD) Requirements defect found through traditional acceptance testing Cost Programming defect found through Continuous Integration Requirements or design defect found through Active Stakeholder Participation Requirements defect found through traditional system testing Programming defect found through traditional system testing Defect found through a review or inspection Minutes Hours Days Months Length of Feedback Cycle Source: Scott Ambler

7 but is also a radical virus that is often rejected by its host VersionOne State of Agile Company philosophy or culture at odds with core agile values Lack of experience with agile methods 42% 46% 63% 44% 41% 47% Lack of management support 38% 38% 45%

8 Vastly different paradigms and controls Mechanistic/reductionist Artisanal/responsive

9 Defined Process Control Any wood used in the ground (i.e. foundations, fence posts) must be H4 treated Spacing of foundations must meet standard, joists must be fixed to foundations using z- nails Foundations and other structural elements must meet durability expectations of 50 years

10 Empirical Process Control Sprint Retrospective Sprint Review

11 PRINCE2 is about control Directing a project Authorize Initiation Authorize the Project Authorize a Stage Plan or Exception Give ad-hoc Direction Authorize Project Closure Issues and Risks Controlling a Stage Highlight Reports Functional Mgr(s) Take Corrective Action Escalate Issues and Risks Review the Stage Status Report Highlights Capture and Examine Issues and Risks Authorize Work Packages Review Work Package Status Receive Completed Work Packages Project Mgr Authority to deliver a Work Package Checkpoint Reports Completed Work Package Accept a Work Package Execute a Work Package Deliver a Work Package Managing Product Delivery Team(s)

12 Agile has very different engagement model WARNING Radical transparency

13 Scrum Do we really not need a PM? Functional Mgr Vision Roadmap Overloaded! Direction Stakeholder Mgmt Control Project Mgr Priorities Remove blockers Product Owner Direction Control Delivery Team(s) Task Mgmt Assist Investor Things missing! Scrum Master Budgeting Resource Planning Contract Mgmt Reporting

14 A choice

15 Managing a Stage Boundary Managing a Stage Boundary Closing a Project Agile is workflow management Release Backlog Backlog Item Item Item SCRUM Backlog Item Item Item Product Backlog Release Plan Sprint Plan Release Backlog Sprint Plan Backlog Item Item Item Backlog Item Item Item Product Backlog Sprint Plan Release Plan Sprint Plan Sprint Plan Daily Scrums Sprint Plan replaces Team Plan for technical work MPD Team Plan(s) MPD Team Plan(s) PRINCE2 Processes & Artefacts SU Project Brief PID + Stage Plan IP CS Updated PID + New Stage Plan CS PID + Stage Plan Closure Report Directing a Project

16 Agile as delivery AND governance Directing a project Lean Start-up Google Design Sprint PoCs/Spikes Authorize Initiation Authorize the Project Issues and Risks Authorize a Stage Plan or Exception Controlling a Stage Give ad-hoc Direction Highlight Reports Authorize Project Closure BAU backlog Backlog prioritisation Sprint Planning Sprint Review Take Corrective Action Escalate Issues and Risks Review the Stage Status Report Highlights Capture and Examine Issues and Risks Authorize Work Packages Review Work Package Status Receive Completed Work Packages Sprint Review Sprint Planning Authority to deliver a Work Package Checkpoint Reports Completed Work Package Accept a Work Package Execute a Work Package Deliver a Work Package Sprint execution Managing Product Delivery

17 Risk Impact So why choose Agile? Analysis Design Code Integrate & System Test Customer Use! First build and deliver Delayed customer feedback and late delivery of value. Highest risk activities such as integration, system testing, load testing are tackled late. Time Adapted from: Craig Larman, Agile & Iterative Development, 2004

18 Risk Impact So why choose Agile? First build and deliver All activities are tackled early. Quick feedback and customer gets early value. Customer Use Customer Use Customer Use Customer Use Integrate & System Test Integrate & System Test Integrate & System Test Integrate & System Test Code Code Code Code Design Design Design Design Analysis Analysis Analysis Analysis Time Adapted from: Craig Larman, Agile & Iterative Development, 2004

19 And you avoid this Picture from Franck Nijimbere Early Delivery so much optimism! Issues. The pressure rises Overtime starts, quality suffers The death march ends

20 Key principle being in control is making explicit trade-offs Sustainable Pace Fixed Scope Time Cost Variable Quality Traditional Timeboxed Quality Variable, but transparent Variable Time Cost Scope Source: DSDM Consortium

21 Key principle MVP is the least you can do and be successful MVP Must Haves

22 New Tools Source: Adam Sherez - Unsplash.com

23 Features Agreeing baselines on a Fixed Time Project Ideal case as good as or better than green line Scope Contingency Deliver whatever is ready at this time Worst allowable case up to red line Time/Cost Musts <= 60%

24 Story Mapping Multiple Releases Join up and log in Search Sprint 1 Sprint 2 Post an auction Set the auction end date Set a reserve Accept bids Sprint 3 Login using gmail Save favourite sellers Sprint 4 Set a buy-now price Auto auction end date of today + 7 days Sprint 5 Determine an auction winner Address verification Chat between buyer and seller Extend an inprogress auction Change the reserve Notify the auction winner by Sprint 6 Book a courier Release 1 Release 2 Release 3 Adapted from work by Jeff Patton

25 Precise enough planning 1.8 metres high 5 metres high 180 metres high

26 WIP, Flow, and Team Composition Backlog Analysis Dev Test Done E F D C B A SME Analyst Dev Dev Tester

27 New governance metrics Working Software Source : Evan Dennis - Unsplash.com Lead time Cycle time Average Velocity

28 Lastly, focus your attention on the right things monthly quarterly 6-12 monthly quarterly Pool of new ideas Idea Triage Portfolio Meeting Budget Meeting Portfolio Meeting Outline Project Brief Detailed Project Brief PID Release Acceptance Integration Agile Process quarterly monthly Product Backlog Adapted from work by Klaus Leopold

29 Wrap-up Small is beautiful, and much less risky Project management and Agile come from different paradigms, and we need skills from both To be really agile, projects need to shift from a push model to a pull model. New levers are scope and quality You need to understand the purpose of events or artefacts and the implications of using/not using them Don t accept proxy measures of progress

30

31 Questions? You can also me: Source : Morvanic Lee- Unsplash.com

32 Injecting fresh thinking to solve tough business problems.