Mattias Skarin Kanban / Lean coach

www.crisp.se

What would you do?

Oct 3, 2011

http://blog.crisp.se/mattiasskarin mattias.skarin@crisp.se

- Learning from charts

Group into 5-8p Choose a team name!

Say Hi to your neighbour!

2

Learning objectives

Mattias Skarin 2

Understanding basics of Control charts, continuous flow chart Put yourself into real shoes – what should be happening Can you beat the monkey?

http://rainbowwallpaper.blogspot.com/2011/04/funny-monkey-cartoon-pics-monkey-funny.html

3

Contributions

Ismael Héry and Benoit Guillou

Henrik Kniberg

Mattias Skarin 3

2011-10-06

4

(Some) valid purposes for collecting data

Mattias Skarin 4

All tools needs a purpose. Know ”why” helps avoid expensive tools

Validating a theory Learning over time

Distinguish between variance and trend

Gain precision

Every learning starts with a question

5

Validating a theory

Mattias Skarin 5

Arrived : Arrived tickets this week (green) Resolved : Resolved tickets this week (black)

6

Validating a theory

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Sprint 1 Sprint 2 Sprint 3 Sprint 4 Sprint 5 Sprint 6 Sprint 7

Value demand

Failure Demand

Average 28 %

Mattias Skarin 6

Data devteam 2009

7

CHARTS BASICS

2011-10-06

Mattias Skarin 7

8

Continuous flow chart

Mattias Skarin 8

Time

Started

Delivered Stories

Cycle time

WIP

9

Control charts

Mattias Skarin 9

Variable

Observation

Average

+ 3σ

- 3σ

10

Goal: separate expected events from unexpected

Mattias Skarin 10

99.7 % of observations occurs within three std dev. of the mean 95 % occurs within two std. dev of the mean 68 % occurs within one std. dev of the mean (Assuming normal distribution)

11

LEARNING FROM CHARTS

Can you beat the monkey?

2011-10-06

Mattias Skarin 11

12

Learning from real cases

There can be multiple solutions to any problem

You are self organizing!

You need to motivate your choice

I get to play god..

Mattias Skarin 12

Thou are allowed to ask questions!

13

Organize

Groups of 5-8

Keep score

Pick a team name

Mattias Skarin 13

14

The case

Mattias Skarin 14

Sprint 1 Sprint 2 Sprint 3

The problem: Why do we always work with 5 projects in parallell although we plan for two?

15

What should be happening?

Mattias Skarin 15

0

5

10

15

20

25

1 2 3 4 5 6 7

To do

In Dev

To test

Done

16 Mattias Skarin 16

0

5

10

15

20

25

1 2 3 4 5 6 7

To do

In Dev

To test

Done

33% Todo (waiting) 17 % Coding 50% Testing

40% Todo (waiting) 50 % Coding 10% Testing

17

What should be happening?

A. Assign a WIP on number of projects

B. Pair program if you get stuck

C. Hold back specification, until just before development

D. Deliver, when testing is complete

E. Other

Mattias Skarin 17

A. Assign a WIP on number of projects

B. Pair program if you get stuck

C. Hold back specification, until just before development

D. Deliver, when testing is complete

E. Other

18 Mattias Skarin 18

More examples exists but for now only demoed live

19

What can trigger change?

Mattias Skarin 19

Questions – someone asking them New ideas – how to do it better Consequence awareness – ”this will happen if change does not take place” Consolidation – a momentum grows large enough to overcome the threshold

Will to experiment – someone willing to give it a try A failure – ”uh-uh that didn’t work”

Gradual

Fast

20

Some final thoughts

Charts + Situation knowledge = learning

Useful in times of stress

Keep it simple. Plot on your whiteboard.

Not all facts trigger change

Human action is required

Mattias Skarin 20

21

Thank you!

Mattias Skarin 21

”Change is not necessary. Survival is optional” - W. E Deming