SPSE2012 - Measure and Manage Flow in Practice
4 likes1,023 views
The document summarizes Zsolt Fabok's experience implementing Kanban principles in his project management. It describes four issues they initially faced: having too many work items in progress, long lead times, too much rejected work, and variability in work item size. Through categorizing work items and setting service level agreements, they improved flow by limiting work in progress, reducing lead times and rejection rates, and balancing the portfolio. The cumulative flow diagram and statistics on lead times and work hours illustrate how visualizing and measuring flow helped manage their project progress.
SPSE2012 - Measure and Manage Flow in Practice
- 1. Measure and Manage Flow in Practice by Zsolt Fabók @ZsoltFabok March 19, 2012
- 2. Let’s start with a game!
- 3. Price: ~67000 HUF (~ 2045 SEK)
- 7. actual progress 90 68 progress 45 23 0 cw12 cw13 cw14 cw15 cw16 cw17
- 8. Decisions must be made based on facts and objective data!
- 9. I went looking for something which provides an accurate and objective view on my projects and makes them predictable!
- 10. And I met Joshua Kerievsky in 2009...
- 11. And I met Joshua Kerievsky in 2009... Kanban is getting popular, you should have a look
- 12. The five core principles of Kanban* * based on David J. Anderson’s Kanban
- 13. The five core principles of Kanban* Q D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ * based on David J. Anderson’s Kanban
- 14. The five core principles of Kanban* Q 3 4 1 2 D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Limit work in progress ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ * based on David J. Anderson’s Kanban
- 15. The five core principles of Kanban* Q 3 4 1 2 D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Limit work in progress ~ ~ ~ ~ Measure and manage flow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ lead time * based on David J. Anderson’s Kanban
- 16. The five core principles of Kanban* Q 3 4 1 2 D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Limit work in progress ~ ~ ~ ~ Measure and manage flow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ cycle time lead time * based on David J. Anderson’s Kanban
- 17. The five core principles of Kanban* Q 3 4 1 2 D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Limit work in progress ~ ~ ~ ~ Measure and manage flow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ cycle time throughput lead time * based on David J. Anderson’s Kanban
- 18. The five core principles of Kanban* Q 3 4 1 2 D Visualize workflow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Limit work in progress ~ ~ ~ ~ Measure and manage flow ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Make process policies explicit Improve collaboratively cycle time throughput lead time * based on David J. Anderson’s Kanban
- 20. The Cumulative Flow Diagram Done number of work items Started Queued time
- 21. The Cumulative Flow Diagram Done number of work items Started Queued lead time time
- 22. The Cumulative Flow Diagram Done number of work items Started Queued cycle time lead time time
- 23. The Cumulative Flow Diagram Done number of work items Started Queued WIP cycle time lead time time
- 24. The Cumulative Flow Diagram Done number of work items Started Queued backlog WIP cycle time lead time time
- 26. Q 3 4 1 2 D ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
- 27. Q 3 4 1 2 D ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
- 28. Q 3 4 1 2 D ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ + ~ ~
- 29. Q 3 4 1 2 D ~ ~ ~ ~ ~ ~ Simplicity ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ + = Objectivity ~ ~ Predictability
- 30. Enough with the boring theoretical stuff!
- 31. The history of a real project Done Started Queued The collected data is the courtesy of Digital Natives
- 33. #1: Too many work items in progress Done Started Queued
- 34. We’ll wait for you with testing... No worries!
- 35. #2: Way too long lead time Done Started Queued
- 37. Distribution of lead times count 15 13 10 8 5 average 3 median* 0 days 1 3 5 7 9 11 13 15 22 33 *Calculation of medians is a popular technique in summary statistics and summarizing statistical data, since it is simple to understand and easy to calculate, while also giving a measure that is more robust in the presence of outlier values than is the mean. Wikipedia
- 38. Some examples of work items with 8-day lead time time spent on time spent waiting implementation ID description # % (hours) (hours) 1 63 98 7 57 90 2 62 97 2 62 97 3 61 96
- 39. 95% waste
- 40. Distribution of lead times Before count 15 13 10 8 5 3 0 days 1 3 5 7 9 11 13 15 22 33 average median
- 41. Distribution of lead times Before After count count 15 15 13 13 10 10 8 8 5 5 3 3 0 0 days 1 3 5 7 9 11 13 15 22 33 days 1 3 5 7 9 11 13 22 average median
- 42. We are soooo fast!!!
- 43. #3: Still too much work in progress Done Started Queued
- 44. Number of rejected work items count 15 11 8 4 0 week 31-32 33-34 35-36 37-38
- 45. Number of rejected work items count 15 11 8 4 0 week 31-32 33-34 35-36 37-38 39-40 42-43 44-45
- 47. #4: Improving by categorizing work items Done Started Queued
- 48. All the work items we had so far (~20 work items) v ~ v ~ ~ ~ v ~ v ~ v ~ v v ~ ~ ~ ~ ~ v
- 49. Categorizing them into three groups S v ~ ~ v ~ ~ ~ M ~ ~ ~ v v ~ ~ v v L ~ v ~ v
- 50. The lead time distribution count 4 3 3 ~ ~ ~ 2 M ~ ~ ~ v v ~ ~ v v 1 1 0 days 1 2 3 4 5 6 7 8 9 12 13 16
- 51. The lead time distribution count 4 3 3 ~ ~ ~ 2 M ~ ~ ~ v v ~ ~ v v 1 1 0 days 1 2 3 4 5 6 7 8 9 12 13 16
- 52. The lead time distribution count 4 3 3 ~ ~ ~ 2 M ~ ~ ~ v v ~ ~ v v 1 1 0 days 1 2 3 4 5 6 7 8 9 12 13 16 SLA
- 53. The spent time distribution count 6 5 4 ~ ~ ~ 3 M ~ ~ ~ v v ~ ~ v v 2 1 0 hours 6 7 8 9 10
- 54. The spent time distribution count 6 5 4 ~ ~ ~ 3 M ~ ~ ~ v v ~ ~ v v 2 1 0 hours 6 7 8 9 10 SLA
- 61. Thank you very much for your attention! http://zsoltfabok.com/ zsolt.fabok@ericsson.com