Why Do We Learn More from What Did Not Work?

By Pascal Dennis (bio)

Building on Al's recent blog, why do we learn from more failure than success?

Seems to me, it's because failure illuminates more of the design space than success.

Supposing we're testing the structural integrity of say, a hard hat, by dropping a heavy weight on it.

If we test to the standard, (say 20 kg) and the hard hat remains intact, you've learned something about what sort of blow it can sustain.

But suppose we keep dropping heavier & heavier weights, and vary the angle of the blows - until the hard hat shatters.

Our analysis of the fragments, breakage pattern, of the slow motion video and so on, will teach us far more about the nature of hard hats.

That's why experienced labs & design teams test to failure.

A caveat, as Al suggests, is that we fail quick & fail often, (to minimize hassle & transaction cost.)

A second caveat: our failures are controlled & buffered so nobody gets hurt!

These same principles apply in strategy, product & process design and problem solving.

That's why we say 'problems are gold'.

We have to be comfortable, of course, with experimentation & ambiguity.

In my experience, the best leaders create a sense of free-wheeling energy & opportunity.

"Let's try some stuff -- and see what happens!"

"Holy cow, who would have thought...!?"

Best,

Pascal


In case you missed our last few blogs... please feel free to have another look…

Failure is a Requirement for Innovation
KAIZEN – Small Changes vs. Monster Projects
Is Inventory a waste or a cover-up of deeper waste?
7 Basic Quality Tools – Are they underrated?

Failure is a Requirement for Innovation

By Al Norval (bio)

As the parent of two children, I often found myself trying to protect my kids from failure. As I reflect back on that now, I’m not sure I always made the right decisions in doing so. How could I expect my kids to learn if they never experienced failure. Even more difficult, how do I get them to be more comfortable with failure. Anyway, more on the dilemmas of parenting in another blog.

The same principle holds true in Lean as we apply the Scientific Method to problem solving. In the scientific method we set out a hypothesis and then run experiments to test our thinking. The test needs to be binary – it either succeeds in which case we implement and standardize our countermeasure or it fails in which case we need to develop another countermeasure.

The key to the scientific method is that we learn something in both cases which is why it is critical for organizational learning. Not only do we learn when we succeed but we learn how not to do things when we fail. In fact, many times success only comes after repeated failure. Thomas Edison, one of the greatest inventors of all time, is a great example of someone who embodied this principle.

But what happens if we apply the scientific method, develop a hypothesis and run an experiment on things we already know and understand?

Result – we can follow the methodology but no organizational learning occurs.

To learn we need to run experiments and test things where we don’t know the outcome. This leads to the mantra:

“Experiment, Fail, Experiment again, Fail Again but this time Fail better”

Failing better means we’re learning. We’re learning what doesn’t work which means we’re closer to learning what will work. This learning allows us to develop more creative and innovative countermeasures to address the problem. Combine this with “Fail Fast” and we have the formula for rapid innovation within organizations. As a side note, “Failing Fast” doesn’t necessarily help an organization if they aren’t learning from their failures.

For leaders this means becoming comfortable with failures, encouraging teams to try new things and ensuring the salient learning points are captured from the experiments.

Doing so will help ensure organizational success and wise leaders know, it’s only a thin line that separates success from failure.

Cheers

Al


In case you missed our last few blogs... please feel free to have another look…

KAIZEN – Small Changes vs. Monster Projects
Is Inventory a waste or a cover-up of deeper waste?
7 Basic Quality Tools – Are they underrated?
What Does Leader as a Teacher Really Mean?

Why Do We Learn More from Failure?

By Pascal Dennis

Why do we learn from more failure than success?

Seems to me, it's because failure illuminates more of the design space than success.

Supposing we're testing the structural integrity of say, a hard hat, by dropping a heavy weight on it.

If we test to the standard, (say 20 kg) and the hard hat remains intact, you've learned something about what sort of blow it can sustain.

But suppose we keep dropping heavier & heavier weights, and vary the angle of the blows - until the hard hat shatters.

Our analysis of the fragments, breakage pattern, of the slow motion video and so on, will teach us far more about the nature of hard hats.

That's why experienced labs & design teams test to failure.

A caveat, as my colleague, Al Norval, suggests, is that we fail quick & fail often, so as to minimize hassle & transaction cost.

A second caveat: our failures are controlled & buffered so nobody gets hurt!

These same principles apply in strategy, product & process design and problem solving.

That's why we say 'problems are gold'.

We have to be comfortable, of course, with experimentation & ambiguity. Which means we need a strong foundation built on the fundamentals.

The fundamentals – Value & Waste, Standardized Work, Visual Management etc. – anchor us, so we feel comfortable with the inherent instability of rapid experimentation.

In my experience, the best leaders teach the fundamentals, then create a sense of free-wheeling energy & opportunity.

"Let's try some stuff -- and see what happens!"

"Holy cow, who would have thought...!?"

Best,

Pascal

Pascal Dennis Bio

Why Do We Learn More from What Did Not Work?

By Pascal Dennis

Building on Al's recent blog, why do we learn from more failure than success?

Seems to me, it's because failure illuminates more of the design space than success.

Supposing we're testing the structural integrity of say, a hard hat, by dropping a heavy weight on it.

If we test to the standard, (say 20 kg) and the hard hat remains intact, you've learned something about what sort of blow it can sustain.

But suppose we keep dropping heavier & heavier weights, and vary the angle of the blows - until the hard hat shatters.

Our analysis of the fragments, breakage pattern, of the slow motion video and so on, will teach us far more about the nature of hard hats.

That's why experienced labs & design teams test to failure.

A caveat, as Al suggests, is that we fail quick & fail often, (to minimize hassle & transaction cost.)

A second caveat: our failures are controlled & buffered so nobody gets hurt!

These same principles apply in strategy, product & process design and problem solving.

That's why we say 'problems are gold'.

We have to be comfortable, of course, with experimentation & ambiguity.

In my experience, the best leaders create a sense of free-wheeling energy & opportunity.

"Let's try some stuff -- and see what happens!"

"Holy cow, who would have thought...!?"

Best,

Pascal

Failure is a Requirement for Innovation

By Al Norval

As the parent of two children, I often found myself trying to protect my kids from failure. As I reflect back on that now, I’m not sure I always made the right decisions in doing so. How could I expect my kids to learn if they never experienced failure. Even more difficult, how do I get them to be more comfortable with failure. Anyway, more on the dilemmas of parenting in another blog.

The same principle holds true in Lean as we apply the Scientific Method to problem solving. In the scientific method we set out a hypothesis and then run experiments to test our thinking. The test needs to be binary – it either succeeds in which case we implement and standardize our countermeasure or it fails in which case we need to develop another countermeasure.

The key to the scientific method is that we learn something in both cases which is why it is critical for organizational learning. Not only do we learn when we succeed but we learn how not to do things when we fail. In fact, many times success only comes after repeated failure. Thomas Edison, one of the greatest inventors of all time, is a great example of someone who embodied this principle.

But what happens if we apply the scientific method, develop a hypothesis and run an experiment on things we already know and understand?

Result – we can follow the methodology but no organizational learning occurs.

To learn we need to run experiments and test things where we don’t know the outcome. This leads to the mantra:

“Experiment, Fail, Experiment again, Fail Again but this time Fail better”

Failing better means we’re learning. We’re learning what doesn’t work which means we’re closer to learning what will work. This learning allows us to develop more creative and innovative countermeasures to address the problem. Combine this with “Fail Fast” and we have the formula for rapid innovation within organizations. As a side note, “Failing Fast” doesn’t necessarily help an organization if they aren’t learning from their failures.

For leaders this means becoming comfortable with failures, encouraging teams to try new things and ensuring the salient learning points are captured from the experiments.

Doing so will help ensure organizational success and wise leaders know, it’s only a thin line that separates success from failure.

Cheers

Fail Forward to Success

By Al Norval

I watch as organizations practice the PDCA cycle. Typically, they do a good job setting up their plans with a hypothesis and begin to deploy them. But as time passes, they discover things aren’t going according to plan. They get frustrated, angry and jump up and down.

The problem is they expect things to go according to plan and in truth, nothing goes according to plan. That’s one of the reasons why the Check/ Adjust part of PDCA exists and why it is so important. If things don’t go according to plan – problem solve and get things back on plan. Lean organizations understand this and get good at rapid PDCA cycles so they can surface and solve problems quickly and make many, small adjustments to keep them going forward towards their goals.

They also understand the other part of Check/ Adjust. When things do go according to plan, standardize and lock in the process.

In both cases, lean organizations learn from using the PDCA cycle. When things fail the hypothesis at the check step, they problem solve and get to root cause. They learn from this and develop a deeper understanding of what is actually happening. When things pass the hypothesis, they learn as well. The key is they expect to have failures and expect to learn from their failures.

In both cases, it’s important to lock in the learning and to share it across the entire organization. This sets up the formation of a learning organization, one that is based on learning from problem solving using the scientific method.

The faster they check/ adjust, the quicker they learn and the faster they can meet their objectives. Thus Lean organizations get good at using rapid experimentation and many small PDCA loops.

This concept was summed up nicely in the following quote:

“Virtually nothing comes out right the first time. Failures, repeated failures, are finger posts on the road to achievement. The only time you don’t want to fail is the last time you try something. One fails forward to success”

– Charles F. Kettering

The only real failure is not learning from our failures.

Cheers