Planning for Value
4 likes1,335 views
The document outlines strategies for agile value-based planning in software development, emphasizing the importance of understanding stakeholder needs and valuing work based on delivery timelines. It discusses various methodologies for estimating return on investment (ROI) and managing project risks while prioritizing features to maximize revenue. Additionally, it highlights the critical role of effective communication, feedback loops, and phased deliveries in achieving project success and addressing business expectations.
Planning for Value
- 1. Planning for Value Value Based Planning A #NoProjects production Allan kelly @allankellynet allan@allankelly.net http://www.allankelly.net
- 2. Allan Kelly… Training for Agile Advice on Agile, adopting Agile Organizing teams Author – Xanpan: Team Centric Agile Software Development https://leanpub.com/xanpan (2014-2015) – Business Patterns for Software Developers (2012) – Changing Software Development: Learning to be Agile (2008) allan@allankelly.net http://www.allankelly.net @allankellynet (Twitter)
- 3. Quantity of software Price of software0 Supply (Development) Demand for software Supply with Agile Agile toolkit helps boost supply but has little to address demand
- 4. “The Business” wants to know when something will be done
- 5. Rational end dates? In the US more than half of the large projects … predetermined end date is selected, and it is forced on the project by arbitrary decree. Capers Jones, 2008
- 6. The problem “The Business” wants to know when something will be done it done by a date
- 7. Project Success Criteria • On Schedule • On Budget • On Quality (Features) Where’s the value? £ $ €
- 8. The problem “The Business” wants to know when something will be done it done by a date to make money
- 9. Exercise • Build an online shopping site • Small groups • Imagine you are a software provider • 4 minutes • See story handout • Suggest: approach, time, people, bid
- 10. Value is an input to engineering Solution must be built to: a) Obtain as much of the value as possible b) For a cost which leaves some profit How much profit is made and How much value is extracted are negotiable.
- 11. How do you get to value? • Ask your stakeholders • Do the analysis • Estimate
- 12. Agile folk get it wrong! • Give more opportunities for feedback • Tighter feedback loops • Allow for changing requirements • Improve quality • Reduce risk • They increase value • They make more money £€$ (O, and reduce risk, improve quality, etc. etc.) Regular deliveries are good because…
- 13. Cost-Benefit Return on Investment (ROI) Net Present Value (NPV) Expected costs $1,000,000 Expected revenue $1,250,000 12 months to develop Therefore: ROI = (1,250,000 - 1,000,000) = 25% 1,250,000
- 14. Simplistic ROI using NPV Over 12 months 25% is too simplistic Also need to know • “Risk free” rate – i.e. bank rate • Assume 5% Then NPV = $193,309 19.3% Cost Return Net Jan $ 1,000,000 $ 0 -$ 1,000,000 Feb $ 0 $ 0 $ 0 March $ 0 $ 0 $ 0 April $ 0 $ 0 $ 0 May $ 0 $ 0 $ 0 June $ 0 $ 0 $ 0 July $ 0 $ 0 $ 0 August $ 0 $ 0 $ 0 Sept $ 0 $ 0 $ 0 Oct $ 0 $ 0 $ 0 Nov $ 0 $ 0 $ 0 Dec $ 0 $ 1,250,000 $ 1,250,000 Total: $ 1,000,000 $ 1,250,000 $ 250,000
- 15. Cash gets spent month by month Cost Return Net Jan $ 83,333 -$ 83,333 Feb $ 83,333 -$ 83,333 March $ 83,333 -$ 83,333 April $ 83,333 -$ 83,333 May $ 83,333 -$ 83,333 June $ 83,333 -$ 83,333 July $ 83,333 -$ 83,333 August $ 83,333 -$ 83,333 Sept $ 83,333 -$ 83,333 Oct $ 83,333 -$ 83,333 Nov $ 83,333 -$ 83,333 Dec $ 83,333 $ 1,250,000 $ 1,166,667 Total $ 1,000,000 $ 1,250,000 $ 250,000 Big bang delivery $83,000 per month $1.25m at end Risk free rate 5% NPV = $215,725 21.6%
- 16. Monthly deliveries Cost Return Net Jan $ 83,333 $ 104,167 $ 20,833 Feb $ 83,333 $ 104,167 $ 20,833 March $ 83,333 $ 104,167 $ 20,833 April $ 83,333 $ 104,167 $ 20,833 May $ 83,333 $ 104,167 $ 20,833 June $ 83,333 $ 104,167 $ 20,833 July $ 83,333 $ 104,167 $ 20,833 August $ 83,333 $ 104,167 $ 20,833 Sept $ 83,333 $ 104,167 $ 20,833 Oct $ 83,333 $ 104,167 $ 20,833 Nov $ 83,333 $ 104,167 $ 20,833 Dec $ 83,333 $ 104,167 $ 20,833 $ 1,000,000 $ 1,250,000 $ 250,000 Per month $83,000 per/month $104,167 value/month Risk free rate 5% NPV = $243,359 24.34%
- 17. Properly modeling NPV Increases value by $50,000 Adds 5% to return
- 18. Follow the money Is this project right for Agile? Or is Waterfall better here? Who cares about right? Agile will make more money
- 19. Which story first? Value = $355,000 Time to delivery = 4 weeks Value = $1,060,000 Time to delivery = 6 weeks As a toy retailer I want an app that allows kids to select from my special range so that their parents can buy. As a toy retailer I want an app that allows kids to make lists of toys they want so their parents can buy.
- 20. Which story first? Value = $355,000 Time to delivery = 4 weeks Value = $1,060,000 Time to delivery = 6 weeks As a toy retailer I want an app that allows kids to select from my special HALLOWEEN range so that their parents can buy. As a toy retailer I want an app that allows kids to make lists of toys they want SANTA TO BRING so their parents can buy. Today is 1 September
- 21. Santa app total value over 4 months 0 200000 400000 600000 800000 1000000 1200000 Sep 1 Sep 15 Sep 29 Oct 13 Oct 27 Nov 10 Nov 24 Dec 8 Dec 22 Jan 5 Value over 4 months Delivery date Total value for a given date Total value starts to fall after 15 September To maximize value deliver by 15 Sep1 Sep-13 Oct 6 weeks
- 22. Halloween app value over 4 months 0 50000 100000 150000 200000 250000 300000 350000 400000 Sep 1 Sep 15 Sep 29 Oct 13 Oct 27 Nov 10 Nov 24 Dec 8 Dec 22 Jan 5 Product B value over 4 months
- 23. Together 0 200000 400000 600000 800000 1000000 1200000 1400000 1600000 Sep 1 Sep 15 Sep 29 Oct 13 Oct 27 Nov 10 Nov 24 Dec 8 Dec 22 Jan 5 Combined value from two apps Santa app Halloween app
- 24. Together 0 200000 400000 600000 800000 1000000 1200000 1400000 1600000 1 2 3 4 5 6 7 8 9 10 Combined value from two apps Halloween app Santa app
- 25. What do you want to do? A. Do Halloween & forget Santa B. Do Santa & forget Halloween C. Walk away D. Do both and pray E. Add more people F. Change the estimates G. None of the above
- 26. Santa first App makes $1,025,000 $35,000 lost because it was not ready by 13 October Halloween app makes $0 Total = $1,025,000
- 27. Halloween app makes $340,000 $15,000 lost because not ready before 29 September Santa makes $800,000 $225,000 lost because it is not ready until 10 November Halloween first
- 28. Halloween comes first Santa first Total = $1,025,000 Halloween first Total = $1,140,000 $115,000 more
- 29. Cost of Delay • Extra costs incurred because delivery is late – e.g. temporary staff, priority shipping, penalty fines • Revenue foregone: – Not being in the market (less time on sale) – Not being available before critical date – Not being in the market early (competitors)
- 30. “6 month delay can be worth 33% of life cycle profits” McKinsey (Reinertsen) 1983
- 31. Deadlines are elastic by value We think of deadlines as binary But Usually analogue Different dates result in different values
- 32. Explain this… I need this and I need it YESTERDAY O dear, I don’t have a Tardis but maybe I can get it for you tomorrow? Brilliant, I knew I could count on you If all the value accrued yesterday then it is lost If value still exists then it might be worth doing
- 33. Planning Planning work ahead requires: • Knowing what is wanted • Knowing the value of the requests • Knowing the value over time profile
- 35. 3 Planning Horizons Iteration (Sprint) 2-4 weeks ahead Quarter plan (Release) Next quarter 2-6 Iterations - Max 12 wks (2-6 releases ahead) Roadmap 1-2 years by quarter 2-5 year ahead Iteration Quarter plan Roadmap
- 36. Our unit of work is… • Story • User story • Use case • Ticket As a Role or Persona I want to Do a Something So that Objective • Product Backlog Item • MMF, QoV, etc. • A Blue
- 37. 2 Golden Rules As a Role or Persona I want to Do a Something So that Objective Story should benefit business (Story should have value $s & €s) Story should be small – deliverable in days; max 2 weeks
- 38. As a Role or Persona I want to Do a Something So that Objective EPIC TASK Value but too big to deliver soon Small enough to deliver really soon but lack business value
- 39. Epic-> Story-> Task Very valuable but BIG Used for forward planning Breakdown to stories as they approach Epic Story Task StoryStoryStory Task Task The things you do to build a story Very small but no business benefit Stories have value & add acceptance criteria
- 40. 3 Planning Horizons Iteration Quarter plan Roadmap Roadmap plan with Epics Tasks Stories Epics Iteration plan at Task level Quarter (Release) plan with Stories
- 41. Choose which Epics to breakdown to stories based on value EPIC
- 42. Assign value before effort Only estimate effort after value STORY
- 43. Iteration Team planning Quarter plan – Product Owner Roadmap – Strategic Product Owner Tasks Stories Epics
- 44. 3 Planning Horizons Iteration plan Quarter plan Roadmap Context Operational Tactical Strategic Who Team Product Owner (Tactical) Product Owner (Strategic) Horizon 1 to 4 weeks forward Max 12 weeks forward Three months to many years forward Detail level Task Story Epic / Customer Problems Frequency Every 2 weeks Ongoing, at least every 2 weeks Regular, ad hoc Certainty High Probable Low Reason Action work Prepare work Scenario thinking Identify opportunities Tasks Stories Epics
- 45. Iteration Quarter plan Roadmap Epics only move from Roadmap to Quarter when value is known & they are needed within 12 weeks. When they move, break them down to stories Tasks Stories Epics 1 Stories move from quarter to iteration plan when they are scheduled. Break down to tasks if helpful
- 46. • Know the value/time profile • Know your target profit • Know how much you can spend • Work back to date & burn rate • Now engineer
- 47. The Project problem • You have one team • You have one year • You have two projects – Projects have similar value-time profiles
- 48. 0 50000 100000 150000 200000 250000 300000 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1 Nov 01-Dec Project X - Ideal delivery Everything on 1 January A B C D E F If only we could have everything in Project X on day 1 (Jan 1)
- 49. 0 50000 100000 150000 200000 250000 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1 Nov 01-Dec Project X - Monthly delivery schedule A B C D E F In the Real World™ only 1 thing per month
- 50. 0 50000 100000 150000 200000 250000 300000 1 Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1 Nov Project Y - Ideal delivery Begins 4 months after X U V W X Y Z Project Y is similar but value starts in April
- 51. 0 20000 40000 60000 80000 100000 120000 140000 160000 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1 Nov 01-Dec Project Y - Actual delivery after X U V W X Y Z But… a) Project Y cannot start until June (when X has finished) b) Y only delivers one thing a month too
- 52. The Problem… • Y could deliver lots of value in April But • The team is finishing low value items from X in April
- 53. 0 50000 100000 150000 200000 250000 300000 350000 400000 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1 Nov 01-Dec Why wait for value? A B C D E F U V W X Y Z
- 54. Commercial break
- 55. Managing Beyond Projects #NoProjects day long workshop 4 December https://www.eventbrite.co.uk/e/managing- beyond-projects-noprojects-tickets-529113593 https://goo.gl/Y5DwVQ
- 56. Xanpan Amazon or LeanPub https://leanpub.com/xanpan Print http://tinyurl.com/nnu7hom Questions? allan kelly allan@allankelly.net www.softwarestrategy.co.uk Twitter: @allankellynet
Editor's Notes
- #18: Modeling by month increases value marginally