Zsolt Fabok
October 18, 2012
710

# Measure and Manage Flow in Practice

The sixth iteration of this talk, presented at Lean Kanban France 2012 (the previous iteration was presented at Lean Agile Scotland 2012)

October 18, 2012

## Transcript

1. ### Measure and Manage Flow in Practice @ZsoltFabok http://zsoltfabok.com/ #lkfr12 http://lkfr.org/

by Zsolt Fabok 2012.10.18 Broke the WIP limit TWICE Still on the team
2. ### 5 Stories from the life of a team by using

real application and data The collected data is the courtesy of Digital Natives

4. ### WIP point of observation Visualize the situation with Cumulative Flow

Diagram solved
5. ### The Cumulative Flow Diagram Done Started Queued lead time cycle

time WIP backlog time number of work items It offers more than just the WIP...
6. ### ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Q D 3 4 1 2 The simplest way of collecting data:

detour:

11. ### Let’s add more people to the project so that “things

speed up”! /* detour */

detour */
13. ### Actually, “things slow down”, so it was not a good

idea (solve the right problem instead - systems thinking). End of the detour.

17. ### Distribution of lead times days count 0 3 5 8

10 13 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 22 28 33 56 average median* *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
18. ### time spent on implementation (hours) time spent waiting (hours) %

1 63 98 7 57 90 2 62 97 2 62 97 3 61 96 Some examples of work items with 8-day lead time

20. ### Distribution of lead times days count 0 3 5 8

10 13 15 1 4 7 10 13 16 33 Before average 0 3 5 8 10 13 15 1 3 5 7 9 11 13 22 count days After median

23. ### 0 4 8 11 15 31-32 33-34 35-36 37-38 Number

of rejected work items count week
24. ### 0 4 8 11 15 31-32 33-34 35-36 37-38 39-40

42-43 44-45 Number of rejected work items count week

26. ### Sales: “I want to know when the new features can

hit the market!” Management: “I want to know how much it will cost me!”
27. ### All the work items we had so far (~20 work

items) ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ v v v v v v v v
28. ### Categorizing them into three groups ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ v v v v v v v v S M L
29. ### The lead time distribution ~ ~ ~ ~ ~ ~

~ ~ v v v v M 0 1 1 2 3 3 4 1 2 3 4 5 6 7 8 9 12 13 16 days count
30. ### The lead time distribution ~ ~ ~ ~ ~ ~

~ ~ v v v v M 0 1 1 2 3 3 4 1 2 3 4 5 6 7 8 9 12 13 16 SLA days count
31. ### The lead time distribution ~ ~ ~ ~ ~ ~

~ ~ v v v v M 0 1 1 2 3 3 4 1 2 3 4 5 6 7 8 9 12 13 16 SLA days count Expired
32. ### The spent time distribution ~ ~ ~ ~ ~ ~

~ ~ v v v v M 0 1 2 3 4 5 6 6 7 8 9 10 hours count
33. ### The spent time distribution ~ ~ ~ ~ ~ ~

~ ~ v v v v M 0 1 2 3 4 5 6 6 7 8 9 10 SLA hours count

35. ### #1 We decided that we would force ourselves to keep

the SLA #2 Nothing changed. Still the same ratio

38. ### “If you can not measure it, you can not improve

it.” Lord Kelvin http://en.wikipedia.org/wiki/File:Lord_Kelvin_photograph.jpg
39. ### “If you start measuring something you start optimizing it, and

I know it's the wrong thing to optimize.” Paul Graham http://paulgraham.com/swan.html

42. ### 1. We develop software not models (value) 2. Demand ﬁrst,

supply second 3. Observe the system (lead time, throughput) 4. Start measuring, look back if necessary 5. Manage 6. Mind that data expires 7. Goto step 3.