system dynamics seminar, cambridge, ma; april 5, 2002 innovation implementation: how the how matters...
TRANSCRIPT
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
INNOVATION IMPLEMENTATION:HOW THE HOW MATTERS
J. Bradley Morrison
System Dynamics GroupAlfred P. Sloan School of Management, MIT
30 Wadsworth St., Cambridge, 02142voice:617/253-8094 fax:617/258-7579
e-mail:<[email protected]>
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Motivation
· Why is it so hard for people in organizations to make happen what they want to happen?
· Innovation implementation has a poor track record.– May be due to failure of innovation– May be due to failure of implementation
· Usual question: what distinguishes implementation success from implementation failure?– Practices associated with success versus failure
· How to policies for managing implementation affect the process?
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Motivation
· Approaches to enhancing organizational capability through participatory process improvement programs are widespread. For example:– Total Quality Management (TQM)– Total Productive Maintenance (TPM)– Lean Manufacturing (TPS)– High Performance Work Organizations (HPWO)
· Common characteristic that they rely on the active involvement of and contribution from front-line employees to generate improvements.
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
“Start and Fizzle”(Reference Mode)· After a few months, management and shop workers
proudly point to the initial success of the effort.· “The people that were there, they seemed very excited about it.
When they saw the results, and what could happen, got a visual look, basically, of how we're doing things today and how it can actually be, they got kind of excited about it.”
· “They got people looking at the machines, finally fixing them. We got the layout running. You got a pull system in place. And - start looking at some of the overall numbers - they're outstanding. Scrap has come down [thousands of dollars]. Performance went
up from 70% to 94%.”
· Several months later, some new work practices had been abandoned and performance had deteriorated.
· “The wheels are coming off.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Basic Stock and Flow
Tasks ToDoIdeas
GeneratedDoing Tasks
ProcessCapability
ProcessImprovements
Avg Improvementper Task
ProcessLeakage
NormalLeakage
Max Leakage
Min Time toLeak
Normal IdeaGeneration
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· The intent was to learn together as the initiative progressed:
· “Do we just go down there and say - ok folks, we're going to implement lean and we're the experts here? We're going to tell you what to do and you just trust us. ... Do what we tell you and we'll be ok - all right? And we said - no, we can't do that. ... Well, what if we all learn this stuff together, you know? ... Learn a little, do a little. Let's teach these folks a little bit. Let's learn a little bit about this
stuff ourselves. Let's put our heads together.”· But in practice, the work of implementing fell largely
on the shoulders of complementary resources.
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· Workers generate ideas for improvement, but these create task demands on complementary resources:– Engineering– Maintenance– Materials handling– Work group advisors
· “So we went up there and did the analysis, got some of the tools ordered, and did some of the basic things. Then we were starting to require more and more engineering support because this is where the timing came in. That's where it really fell down. I think probably the biggest thing we were missing on that team is we had no one to go to.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· Existing norms about who does what guided the assignment of tasks to specific resources:
· Engineers knew what they had to do:“When you are sitting in a group and it is some work group members and it is work group advisors … and you are talking about how you are going to change this process and you are talking about moving machines around and talking about changing the manual work … You knew what your part of that job was. You had to do the process documentation. You were responsible for moving the equipment and making the layouts and stuff.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· Availability of necessary resources was limited:· “If we had to have something from maintenance, they said you have to
put a [requisition]. It took a month to put lines on the floor.”· “Our space is being cluttered. ... We'd get promised, oh, by tomorrow
that stuff will be out of there. A week later it will be sitting there. Who's responsible? Whose stuff is this?.”
· “How do you do that? Send in a work order. Well, you send in a work order and it disappears. How do you get the priority? Because we are trying to show some speed and show some commitment, but we didn't know who to plug into.”
· “It wasn't really [anything] overt that we're not going to do this and not going to do that, but it was just like [pause] you're pulling an ox cart through a mud pit is basically what you felt like.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· So, tasks to do kept piling up:· “So we just kept on plotting on with our stuff. Had the people creating
things. I think we got the workgroup so far ahead of engineering that - it was bad, but there was no way that I was going to wait for a support organization to give me resources while I got the people engaged. You can't. It's like - they have to catch up.”
· “You know we hadn't finished the first new improvement. … You have all of these things but you didn't get anything accomplished because you have so much you have to do. We are trying to do it all in one time span. … getting all of that stuff done, you have to work with one thing at a time and that was the missing piece to the puzzle. … People kept coming in with more inputs and that was before you had one output.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Two Ways to Get Things Done
Tasks ToDoIdeas
GeneratedDoingTasks
Backlog ofWork
SchedulePressure
Fraction of TimeCollaborating
Tasks DoneCollaboratively
Time SpentCollaborating
Time SpentDoing Tasks
Tasks DoneOtherwise
+
++
+
+
+
-
-ResourcesAvailable
-
+
-
B
WorkingAlone R
WorkingCollaboratively
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· What could be done about the backlog piling up:· “Nothing. You have to wait. That is pretty much it. You could have
a meeting. Well, … we had this action item review every week, and [they] said here is the list of stuff that we want to get done, whatever resources involved and … how much time it is going to take. … It was just a follow-up. What we'd do is try to push the pencil on some of the issues.”
· Or, improvise solutions to the resource challenge:– “I’ll Do It Myself”– Cherry-picking– Acting as a Crutch
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· “I’ll Do It Myself”· “Still no lathe. Still no standard for the lathe. This thing was
supposed to be in there in December. We don't know how long it takes to do this and to do that. I can estimate it, but I can't estimate it in [our] system [laughter]. …Then it was the first couple weeks in January. I just got so frustrated. I finally took their standards and I created a layout for them, and I created a work path for them as options.”
· Cherry Picking:– “You do the easy things first.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Getting Things Done
· Acting as a Crutch· “We had a workgroup advisor who was pulling it all together at the
time. … She would check to make sure. Instead of ordering herself, she would go into the computer and check it to see if it was ordered and if it wasn't, she would go down on the floor and tell the people that they were not doing their job. She wouldn't do it but she was always be check the balance person of it. So she would go and make sure.”
· “She's very persistent. She does not give up easily. She's bound and determined that she will succeed one way or another. And she does it without really alienating a lot of people.”
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Model Structure
Tasks ToDoIdeas
GeneratedDoing Tasks
ResourcesAvailable
IndicatedResources
Desired Time toComplete Tasks
SchedulePressure
Fraction ofTime Spent
Collaborating
Time SpentCollaborating
Time SpentDoing Tasks
Backlog ofWork
Tasks DoneCollaboratively
Tasks DoneOtherwise
Normal RelativeProductivity
IndicatedCollaboration
Fraction
Fraction ofCollaborative
Tasks
Min Time to DoTask
Max Rate ofDoing
ProcessCapability
ProcessImprovements
Avg Improvementper Task
NormalImprovement per
Task
ProcessLeakage
NormalLeakage
Max Leakage
Min Time toLeak
Effect of Experienceon Improvement per
Task
Experiencefrom
Collaboration
Relative Fractionof Experience
NormalProductivity
Time to GainExperience
Normal IdeaGeneration
<NormalProductivity>
Total TasksPossible
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Starting the program: pulse ideas and add resources
Ideas Generated
60
200 10 20 30 40 50 60 70 80 90 100
Time (Week)
Ideas Generated : res20 tasks/Week
Resources Available
400
2000 10 20 30 40 50 60 70 80 90 100
Time (Week)
Resources Available : res20 hours/Week
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
But with temporary increases in resources, Process Capability may improve or decline:
Process Capability
600
450
300
150
0
0 10 20 30 40 50 60 70 80 90 100Time (Week)
Process Capability : res20 widgets/WeekProcess Capability : res16 widgets/WeekProcess Capability : res12 widgets/Week
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Early withdrawal does not allow collaboration to build.
Fraction of Time Spent Collaborating
0.6
0.45
0.3
0.15
0
0 10 20 30 40 50 60 70 80 90 100Time (Week)
Fraction of Time Spent Collaborating : res20 DmnlFraction of Time Spent Collaborating : res16 DmnlFraction of Time Spent Collaborating : res12 Dmnl
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Similarly, a larger batch of new ideas overwhelms the resources and leads to decline.
Process Capability
600
450
300
150
0
0 10 20 30 40 50 60 70 80 90 100Time (Week)
Process Capability : res20 widgets/WeekProcess Capability : largerpulse widgets/Week
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Similarly, a larger batch of new ideas overwhelms the resources and leads to decline.
Process Capability
600
450
300
150
0
0 10 20 30 40 50 60 70 80 90 100Time (Week)
Process Capability : res20 widgets/WeekProcess Capability : largerpulse widgets/Week
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Adding an Effect on Idea Generation
Tasks ToDoIdeas
GeneratedDoing Tasks
ResourcesAvailable
IndicatedResources
Desired Time toComplete Tasks
SchedulePressure
Fraction ofTime Spent
Collaborating
Time SpentCollaborating
Time SpentDoing Tasks
Backlog ofWork
Tasks DoneCollaboratively
Tasks DoneOtherwise
Normal RelativeProductivity
IndicatedCollaboration
Fraction
Fraction ofCollaborative
Tasks
ProcessCapability
ProcessImprovements
Avg Improvementper Task
ProcessLeakage
Effect of Experienceon Improvement per
Task
Experiencefrom
Collaboration
Relative Fractionof Experience
NormalProductivity
Normal IdeaGeneration
<NormalProductivity>
Total TasksPossible
Effect of Experience onIdea Generation
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
With an effect of the rate of idea generation:
Process Capability
800
600
400
200
0
0 30 60 90 120 150 180 210 240 270 300Time (Week)
Process Capability : varqtyweak widgets/WeekProcess Capability : varqtystrong widgets/Week
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Insights and Implications· Equilibrium condition for Process Capability is: inflow =
outflow– Sustained improvement requires building something else,
perhaps the capacity to improve
· Explicit Recognition of the Stock of Tasks to Do– Manage inflow– Manage outflow– Manage consequences
· Increase productivity - shortcuts– “I’ll Do It Myself”– Cherry-picking– Acting as a Crutch
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Managing A Stock of Tasks to Do
· Increase the Outflow– Increase productivity - really– Increase productivity - shortcuts– Add resources– Cancel ideas
· Decrease the Inflow– Reduce the idea generation– Screen ideas
· Accept the Accumulation– Adjust expected completion time– Mitigate consequences of delays
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
4
Managing A Stock of Tasks to Do
Increase the Outflow· Increase productivity - really
– Skill development through experience over time– Training in new tools and techniques– Technology enhancements
· Increase productivity - shortcuts– “I’ll Do It Myself”– Cherry-picking– Acting as a Crutch
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
PORTIA:Good sentences and well pronounced
NERISSA: They would be better, if well followed
PORTIA:If to do were as easy as to know what were good to do, chapels had been churches and poor men's cottages princes' palaces. It is a good divine that follows his own instructions: I can easier teach twenty what were good to be done, than be one of the twenty to follow mine own teaching.
Shakespeare, The Merchant of Venice, Act I Scene 2
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
A Familiar Story: Teach a Person to Fish ...
Time SpentFishing for Them Time Spent
Teaching Them
Their FishingAbility
CatchingFish
-++
Their Hunger
++-+B
FeedingThem
B
TeachingThem
R
DependencyLoop
System Dynamics Seminar, Cambridge, MA; April 5, 2002Innovation Implementation: How the How Matters © 2001 J. Bradley Morrison
Adding resources generates lasting improvements.
Collaboration and Capability
6,000 widgets/Week0.8 Dmnl
3,000 widgets/Week0.4 Dmnl
0 widgets/Week0 Dmnl
0 10 20 30 40 50 60 70 80 90 100Time (Week)
Process Capability : moreresources widgets/WeekFraction of Time Spent Collaborating : moreresources Dmnl