In Chuck Norris We Trust, All Others Bring Data How not to argue and get some data instead Hanno Jarvet
May 12, 2015
In Chuck Norris We Trust, All Others Bring DataHow not to argue and get some data
instead
Hanno Jarvet
www.jarvet.com
Chuck Norris
• Chuck Norris doesn’t call the wrong number. You answer the wrong phone.
• Chuck Norris does not estimate, he knows.
• When Chuck Norris says “done”, then it’s “done”.
www.jarvet.com
Expected results
After the session the participants are better able to:• improve efficiency • build transparency• increase the quality of the output and customer
satisfaction• set strategies for future improvements • create accountabilities
www.jarvet.com
What is Lean?
• Deliver continually increasing customer value
– Expending continually decreasing effort
– By leveraging the time and energy of bright, creative workers
2.
3.4.5.
Is the analysis detailed enough and did it probe deeply enough on the rightissues?Is there evidence of proper five-whys thinking about the true cause?Has cause and effect been demonstrated or linked in some manner?Are all the relevant factors considered (human, machine, material, method,environment, measurement, and so on?
DoConfirmation (Results )1. How will you measure the effectiveness of the countermeasures?2. Does the check item align with the previous goal statement?3. Has actual performance moved line with the goal statement?4. If performance has not improved, then why? What was missed?
CheckFollow-up (Actions)1. What is necessary to prevent recurrence of the problem?2. What remains to be accomplished?3. What other parts of the organization need to be informed of this result?4. How will this be standardized and communicated?
Adjust
Background1. Is the topic relevant to the organization's objectives2. Is there any other reason for working on this topic (e.g., learning)? P
Current Condition1. Is the current condition clear and logically depicted in a visual manner?2. How could the current condition be made more clear for the audience?3. Is the current condition depiction framing a problem or situation to be
resolved?4. What is the actual problem in the current condition?5.
Are the facts of the situation clear, or are there just observations and
opinions?
6. Is the problem quantified in some manner or is it too qualitative?
L
A
Goal / Target Condition1. Is there a clear goal or target?2. what, specifically, is to be accomplished?3. How will this goal be measured or evaluated?4. What will improve, by how much, and when?
Root Cause Analysis1. Is the analysis comprehensive at a broad level?2.
3.4.5.
Is the analysis detailed enough and did it probe deeply enough on the rightissues?Is there evidence of proper five-whys thinking about the true cause?Has cause and effect been demonstrated or linked in some way?Are all the relevant factors considered (human, machine, material, method,environment, measurement, and so on?
N
OwnerMentorDate
Countermeasures (Experiments)1.
Is the analysis comprehensive at a broad level?
Theme: Is there a clear theme for the report that reflects the contents?
www.jarvet.com
Problem Solving A3
For boundary-spanning problemsDevelop a Consensus for action
– Boundary –spanning communication• 30 second glance, 10 minutes to read
– Pull based authority• Agreement of those affected by the change
– Owner Responsibility• Team collaboration
– Cautions• Define the problem carefully• Find REAL root cause• Manager as mentor
Sorbek & Smalley: Understanding A3 Thinking
Scopes of A3s
Strategic (6-12 months)
System(1 week – 6 months)
Process(1 week)
The Deming Cycle
Typical PDCA• Plan quickly
– Address Symptoms
• Do immediately– Jump to conclusions
• Check roughly– Act pretty much the
way you did before
High Velocity Organization PDCA
• Plan deeply– Discuss actual situation and
target with everyone affected– Really understand/model the
problem and its root cause
• Do many quick experiments– Validate your thinking– Check implications carefully
• Act systematically– Update and deploy standards
and checklist disciplines
www.jarvet.com
Theme and Background –Look Very Carefully
• Theme is A3 Title– Identifies relevance– Revised as situation understanding improves
• Background identifies problems impact– Why this problem matters
• Impact of problem on organization• Specific and Quantitative –use graphs, tables, etc.
– People affected understand, agree on, and care
• 10 second rule– Reader can assess relevance of A3 within 10 seconds
www.jarvet.com
Current Condition –Ask what we already know
• Specific, detailed, quantitative, concise– Tables, graphs, histograms, value-stream maps, diagrams– Countermeasures (Experiments)– Highlight exactly where problem occurs– Baseline to compare to metrics after countermeasures are
applied
• Engage everyone affected by or causing symptoms– Build Consensus on what is
• Symptoms / Undesirable Effects everyone can see• Foundation of authority to experiment with countermeasures
• Update as understanding improves
www.jarvet.com
Goal –Next Target ConditionModel what we expect
• What baseline change is wanted?– What does perfect look like?– What does the organization need?
• Mentor ensures that the owner has both– Plausible Hypothesis
• Based on best available model/understanding of how the system should work
– Consensus among stakeholders• Target is attainable and desirable
• Update as root cause and countermeasures developed
www.jarvet.com
Root Cause –Model Cause & Effect
• Identify underlying problem(s) causing symptoms – Root cause is typically faulty thinking or assumptions
• Addressing the root cause(s) improves all levels of symptoms/undesirable effects/visible damage.
• Build consensus among stakeholders– Broad agreement on Cause & Effect network– Reflect best current knowledge about how things work
• Some techniques:• Some techniques:
– 5 Whys -track down the– Ishikawa (fishbone) diagram– Cause-effect diagram– Current Reality Tree
www.jarvet.com
Propose Countermeasures –DO [Many Experiments]
• Identify countermeasures for each candidate root cause– Experiments expected to mitigate underlying problem– Suggested by people involved or A3 owner
• Assess each countermeasure– Discuss with Stakeholders affected– Identify expected changes in meaningful measures from
each countermeasure– Select those with most promise
• (DO) Try each selected countermeasure to get evidence for their effectiveness
www.jarvet.com
Results & Follow-up –Check& Act: Know WHY, not just know
how!• For each countermeasure implemented
– What actually happened
• If different than expected, – Why? – Does the model you used correctly represent what happens?– How do the results improve your knowledge of how to think
about your work?
• What will you monitor to know that the problem remains “solved”?
• What additional problems are revealed after the countermeasures are in place?
www.jarvet.com
Learning Results and Follow-up
• The reliable learning comes from checking the results of your experiments.– If the hypotheses in your root-cause analysis are correct, your
countermeasures should make the situation better by the amount you expected.
– If they do not, either your model is wrong or the countermeasure is not correct or sufficient to address the root-cause and you have to try again.
• Learning only has value if it changes the way you act– Improved workflow, better method, better skills, needed
checklist item, better standards, etc.– This is the ACT part or PDCA, change the way you do this kind
of work.
www.jarvet.com
Ishikawa or Fishbone Diagram
Root Cause Analysis
www.jarvet.com
8 wastes
• Overproduction• Motion• Inventory• Transportation• Waiting• Under-utilized people• Defects• Over-processing
www.jarvet.com
A Fantastic Upcoming Conference
• Less2012.org• 12-14 November, Tallinn• 20 speakers, 4 tracks• Register with promocode AGILEESTONIA for a
50 EUR discount