Home > The Lean Post> To A3 or Not to A3
The Lean Post
Sharing how the world is making things better through lean.

To A3 or Not to A3

by Norbert Majerus
February 9, 2016

To A3 or Not to A3

by Norbert Majerus
February 9, 2016 | Comments (8)

Most engineers like structure, process, and standardized forms; especially as a way of applying order when situations seem to be losing theirs. It is this element of standardization that many find appealing — confidence and comfort amid familiarity. Some people also like “one-size-fits-all” tools but they are not always happy with the fit.

When we first learned and implemented six sigma in Goodyear’s R&D function, I tried to awkwardly force the Define, Measure, Analyze, Improve, Control (DMAIC) process onto any problem I confronted. I see the same thing happening today with the use of A3s, leaving many dissatisfied and abandoning the problem-solving approach. Many people at Goodyear walked away from the DMAIC process, pointing out the waste and misdirection of trying to put every problem on a common DMAIC denominator, and then fell into the same frustrating habits of forcing A3s.

Not every tool is a hammer, and not every problem is a nail. Not every situation warrants the use of an A3.

I learned about A3 thinking and how to complete an A3 by reading John Shook’s Managing to Learn. I think most people familiar with A3s understand the technical process as well as we do, but they still fail to use them effectively. For insights on when and why to use an A3, we looked to the Cynefin framework presented by Dave Snowden and Mary Boone. The two scholars write in A Leader’s Framework for Decision Making, “Good leadership is not a one-size-fits-all proposition.” The same can be said for the use of A3s.

Cynefin basically consists of four cause-and-effect domains that leaders confront and must resolve, each requiring a different type of response. I find the four domains also give direction for when to pull out an A3:

  1. Simple: Decisions are based on best practices, standards, and repeating patterns. An A3 would be a tireless overworking of the problem. “Just do it” is the action in most of these cases, for which answers are already known: standardized work, ethical morals, family values, athletic techniques, etc.
  2. Complicated: This is the realm of the known unknowns. Decisions are based on clear cause-and-effect relationships and expert knowledge, and there can be more than one right answer. You don’t need an A3 to explore these complicated problems and solutions: mathematical relationships in R&D, geologists plotting drilling locations, material scientists developing a new rubber formulations, etc.
  3. Complex: This is the realm of unknown unknowns. Decisions may not yield a single right answer, and things can only be understood in retrospect. Here experimentation is necessary and solutions emerge from the investigative process. The A3 leads one through the process of investigation and experimentation, sets the threshold for an acceptable outcome (the goal), and establishes the actions to get there. You may find, as is often the case in R&D, that one complex problem requires the development of multiple A3s. A leadership temptation with complex problems is to impose order, which often leads to a simple solution that rarely solves the real problem. An A3 forces problem solvers to rigorously explore many potential causes of complex problems: inadequate lab turnaround times, an unexpected consumer response to a new product launch, a sudden drop in customer retention, etc.  
  4. Chaotic: You cannot A3 your way out of chaos. Decisions and actions are intended to reestablish some sense of order and security amid high turbulence and high tension. In business this is comparable to responsive troubleshooting. Quickly stop the conditions for chaos, and then apply the right tools to the simple, complicated, or complex issues behind it. No individual or group can solve chaos, but they can respond and prevent the escalation of chaos: a fire in a manufacturing plant, a terrorist attack, aftermath of a natural disaster, economic collapse, etc.

The A3 process – or the more general process used by Goodyear that I describe in Lean-Driven Innovation – is highly flexible (maybe too flexible), and I could probably squeeze any problem onto an A3 if I put my mind to it. But why would I? Think about the problem domain before you pull out the paper or at least when starting the A3. Based on my A3 coaching experience, struggling with the left side of the A3 (problem background, current condition, and goal) is a clue that the problem is simple or complicated and not really complex and applicable to an A3.

Like most companies, we had some early challenges using A3s, but today A3s are widespread in Goodyear product development and elsewhere. We teach the process and coach engineers, who apply the process to solve real problems, much as Sanderson coached Porter in Managing to Learn. Our engineers use A3s daily, such as when formulating a complex product development iteration (learning cycle); the experiment is run, and the results are documented on the A3.

Our A3s are simple and kept to a single sheet of paper or a whiteboard. I often get requests for our “A3 form,” and I encourage people to find and develop their own. I also tell them to use Shook’s work to guide them through the technical steps, and Snowden and Boone’s to understand when to pull it out. I would strongly advise you to do the same.

The views expressed in this post do not necessarily represent the views or policies of The Lean Enterprise Institute.
Keywords:  A3,  problem solving,  root cause
Was this post... Click all that apply
54 people say YES
70 people say YES
30 people say YES
31 people say YES
Related Posts
8 Comments | Post a Comment
John Shook February 09, 2016
9 People AGREE with this comment

Thanks, Norbert, for contributing this useful post connecting the A3 process with Lean Product and Process Development and Dave Snowden’s powerful Cynefin framework. Highly thought provoking!

I agree with you that the sweet spot for the A3 is Snowden’s Complex realm, which refers to situations that call for experiments to gain deep understanding of the conditions. Solving problems with lean thining begins and ends with Grasping the Situation (GtS) and Snowden’s Complex situations are by definition difficult to grasp. And thanks for pointing out that simple situations, with known cause & effect – Just Do It situations – don’t require and are hindered by an act such as pulling out an A3 in the heat of battle.

As for Snowden’s Complicated and Chaos categories, I suggest consideing that there may be useful application of the A3 process in these areas. For Complicated situations, while the usual sequence of a typical A3 problem-solve may not be required, remember that the purpose of an A3 is as much social as it is technical. So, while the cause & effect relationships in place may be clear enough, there may very well not be agreement on how to tackle the situation as a team, as an organization. For such a situation, the A3 process can be powerful. Still, you make a important point that it’s good to be cautious of overkill, using a formal A3 process when it’s not really necessary.

As for Chaotic situations, I certainly agree that no one should be misled into thinking that pulling out an A3 will lead them out of the wilderness of total chaos. But, an effective A3 dialogue can certainly frame the issues and help an individual or team think through what the heck is going on and possibly frame some experiments or lead to new thinking around the dynamics of the ecosystem that’s unfolding around them. Make sense?     

For that matter, it’s worth noting that an A3 itself doesn’t actually solve any of the four types of situations – it just enables a team or individual to better navigate their way through them.  Durward Sobek and Art Smalley got it right when they referred to “A3 Thinking”. We needn’t to pull out an A3 sized sheet of paper for every situation, but we can always use A3 thinking.    

And thanks for sharing Goodyear’s LPPD story with Lean-Driven Innovation - hope it helps a lot of companies.


Reply »

Tom Paider February 09, 2016
1 Person AGREES with this comment

Thanks for the post Norbert. The issue of when to create an A3 and when not to is certainly a constant topic in my company, Nationwide. We have all kinds of groups, from those that mistakenly think that the more A3s the better, regardless of whether they lead to improvement, to those that are resistant to A3s at all, to all states in between. I've noticed better long term results in the groups that go into overdrive A3 mode though than those that are resistant to doing any. They typically find their footing after a while. Your blog (and John's comment) gives me some things to think about in our application of the A3 process. 

Reply »

Jon Miller February 10, 2016
5 People AGREE with this comment

In none of the 4 situations above does it make sense to "pull out an A3." In all above situations, it does make sense to plan, do, check and act. The need for collaboration, communication and documentation determine whether or not you A3.

Reply »

Peter Fritz February 12, 2016
1 Person AGREES with this comment

Thanks, Norbert! A3's are very useful tools and we use them at 3M all the time. A great platform for communicating and keeping the issues/progress in front of teams and management.

Reply »

Tony Stevens February 19, 2016

Interesting article.  Also agree with Jon's comments.  A3s mirror the PDCA sequence in similar ways to 8D, DMAIC and a host of other structured problem solving methods.  However, I've also experienced the dilemma of when it is most appropriately used - ironically made worse by its popularity.  I'm seeing many variations on the theme with A3 formats for storyboards, status reports, proposals, etc. which, I believe, detracts from the true value and promotes the use of "A3s" where other responses would be more effective.

Thanks Norbert for bringing some needed focus to where A3s best add value.  

Reply »

Steve Duffy April 08, 2016
1 Person AGREES with this reply

Good post.

Trying to implement any problem solving system unless it aready exists is difficult and its success lays as much in culture change as learning to apply a new toolset.

If you think about the current state that confronts you then plot the cultural journey it makes the deployment path easier to understand. For info my most recent efforts to introduce a problem solving system has taken just short of 12 months to embed. That is 12 months with a group of people who had already been exposed to PDCA using A3's. We haven't required an A3 level 8 step PPS to solve the biggest quality problem in the factory but we have applied the principles and utilised a Quality Circle approach with its foundations in Go-Look-See.

There has to be a shift from problem managing to problem solving to release the full power of A3 and I've found that breaking the toolset apart and starting with a simple 5 Why helps those under pressure to focus on the root cause rather than fixing the symptom.

Thanks again Norbert.




Reply »

Anthony DoMoe July 06, 2016
1 Person AGREES with this comment


Thanks for the great post.  You really drive home a common problem in process improvement.  I've been in situations where the executive sponsor insisted on full-blown, statistics-driven DMAIC when a lean-focused approach would have been much more appropriate.  I think that the integration of Lean and Six Sigma left the A3 without adequate representation.  The best tool for the job is sometimes the simplest.

Reply »

Norbert Majerus July 08, 2016
1 Person AGREES with this reply

Thanks Anthony

I specially like your comment "The best tool for the job is sometimes the simplest"

We tried many problem solving tools at Goodyear from 8D to DMAIC but often it was about finding the right problem so the tool would fit.

I think "Managing to Learn/A3" enjoyed such a large acceptance at Goodyear because it is so simle that it fits most situations - we use DMAIC, RCA, ... within the A3 if needed.

The only problem I still encounter is the use of an A3 where an A3 really is not really needed - as described in the post


thanks, n

Reply »

What Problems Do You See?
A3 Thinking Roundup
Dueling Methods: 8D and A3
Please include links as plain text URLs only. Do not copy and paste directly from a web page or other document. Doing so may pick up additional HTML that will not function here.
URLs will be converted to functioning links when your comment is displayed on the site.
Here's an example:
See this article for more details: https://www.lean.org/whatslean