My blog is called Systems Thinking, Lean and Kanban. With the word Lean in the title, and being an active part of the Lean and Kanban community, what am I doing posting an article called The Case Against Value Stream Mapping? Has my recent 22 hour flight to Australia had an affect on my brain?
I have been trained in VSM, have used it, and have found it of value. Its a wonderful tool for helping teams in setting up their Kanban boards; mapping their current Value Stream onto the columns on a board. Alan Shalloway has a great animation of this in action. I also encourage Kanban software tool vendors to have a VSM view in their tool, which will show average delay and process times for a Kanban workflow. In addition when I present at conferences, I reference the usage and results of using VSM in one of my case studies at Lonely Planet. So why this post?
Well, its all to do with how the Value Stream Map is used once it has been created. Lets explore how we learn and how to create change.
Action Science
Chris Argyris, an American business theorist, developed a way of explaining behaviour called Action Science.
In Action Science he describes two simultaneous mental models that make it difficult to create change. The first is our Espoused Theory, which describes the model we say we use to describe how we act (or how we would like others to think we act). The second is our Theory in Use, which is the one we actually use to make decisions.
Mental Models are a way to describe a person’s intuitive perception of the world around them. How we act on that world, our decision rules, are based on our mental models. People always behave consistently with their mental models (theory-in-use) even though they often do not act in accordance with what they say (espoused theory). Within the workplace, we hold generalized beliefs about “what is valued in this organization” and “how things get done around here”. We also hold more specific beliefs about events and people. These beliefs are important because they influence and constrain what we do and don’t do in the workplace.
Single Loop vs Double Loop Learning
Argyris also describes two types of learning. Single-loop learning describes how people learn to adjust their actions in response to natural feedback on the success of those actions in achieving a desired result. They liken this style of learning to a thermostat that adjusts the degree of heating or cooling depending on the temperature of the room. In single loop learning, we may change our decisions, but we leave our underlying mental models and decision rules unchanged. This type of learning solves problems but ignores the question of why the problem arose in the first place (using the previous thermostat example; why is the temperature in the room changing?).
In double-loop learning, we change our underlying mental models and decision rules, which in turn produces new results. Double loop learning uses feedback from past actions to question assumptions underlying current views. Double-loop learning requires not only adjusting one’s actions, but also surfacing, challenging and adjusting the governing variables that are usually taken for granted i.e. our beliefs or “mental maps of reality”.
Double-loop learning adds a powerful dimension to previous experiential learning cycles. In previous models, learning was achieved through reflection on the success (or failure) of your actions. However, in the double-loop model, learning is realized through reflection on the validity and usefulness of your beliefs.
Model I-Inhibiting Double Loop Learning
Argyris tells us that when human beings deal with issues that are embarrassing or threatening, their reasoning and actions conform to a model called Model I. Model I behaviour results in defensive behaviours that block exploring underlying mental models and the resulting maturity that arises. Trying to make change in Model I is difficult because you are dealing with their Espoused Theory. It is neither rewarding nor safe for a person to explore or actually change their mental models and decision rules, so there is a wide gap between their Espoused Theory and their Theory in Use.
Model II-Encouraging Double Loop Learning
Argyris describes a much more productive model that he calls Model II. In Model II, it is safe and rewarding to explore underlying mental models and decision rules. The significant features of Model II include the ability to call upon good quality data and to make inferences. It looks to include the views and experiences of participants rather than seeking to impose a view upon the situation. Theories should be made explicit and tested, positions should be reasoned and open to exploration by others.
The Problem with Value Stream Mapping
With the above in mind, what level of learning will a leader experience when shown a Value Stream Map? Its unlikely they will experience double loop learning. Its more likely that when presented with the map, containing various delay times and process times, that may prove embarrassing or threatening, they will use their current mental models i.e. model 1, and look to blame the workers or blame their managers.
If you have ever watched the ‘Back to the floor’ type TV programmes you will know the problem these kinds of things create, i.e. The boss hears something shocking, goes back into the management factory and issues edicts that make things worse not better. The problem, as always, is two fold. Firstly the boss lacks perspective to understand that a systemic change is required (“This problem has been created by me!”) and secondly is let down by the methods they have learned about how to act.
Top management issuing orders, memos and directives alone is insufficient to change employees’ behaviour. Single-loop learning often leads to organisational malaise resulting in symptoms such as defensiveness, cynicism, hopelessness, evasion, distancing, blaming, and rivalry. There is often a large gap between what leaders publicly “espouse” and the real beliefs that guide their actions. In order to effectively come to grips with new situations, the espoused theories need to be aligned with the theories in use.
Double-loop learning techniques help the organisation members learn together and the organisation change. To do this we need to reveal how the work currently works. We need to reveal the thinking behind the current work design. Value Stream Mapping lists each delay and process time in a flow. It identifies current state and future state and plans for implementation. Waste is identified for reduction. But as John Seddon points out
Managers think that cost is in activity, whereas the counter intuitive truth is that cost is in flow.
Process Mapping
An alternative to Value Stream Maps is Process Mapping. We don’t map the process in a room with brown paper and sticky notes. We don’t ask the managers as they are often too removed from the work to know what is actually going on. We don’t send in a Business Analyst from IT, we don’t talk to proxy customers or talk to a spokesperson for an area. We don’t talk to a few token “doers”. Instead we go into where the work is occurring, with a small team comprised of peers of those who do the work. Why peers? workers tend to tell their colleagues the truth about what’s really going on at work. We also take a leader with the team going into the work to perform the mapping.
As John Seddon states
Studying work as systems reveals counter intuitive truths, challenges to convention, that are more easily understood and accepted if someone sees them for themselves, not if they are told. People need to unlearn and relearn.
Mapping is important because it helps staff to change their perspective on the work.
Firstly it helps them to unlearn what they think is going on in the work. Secondly it helps them to relearn how the work should be designed from the customers’ point of view.
In this regard, process mapping is not simply a means to an end. It is part of the change process.
I will post the mechanics of Process Mapping in a future post, but in conclusion, I suggest it would be more likely for effective change to occur if we go into the work to get knowledge of what is really happening in the work, together with a leader who will experience the same learning as a team and will have the authority to make changes to improve. How we document the process, say in a VSM, is less relevant than the learning acquired.
Agile, Lean, Kanban, and System Thinking 
I don’t understand that straw man that VSM leads to managers blaming people. That’s quite a leap.
If that behavior occurs, that’s a problem with the people not the VSM methodology. Any lean writer or consultant worth their salt will coach managers to look at the whole system and not to blame.
Those same people with a process mapping methodology are just as likely to blame, eh?
I think the title of your post should have been “The Case Against Managers Acting Like Jerks and Blaming People.” I think that’s where your criticism should be placed, not against Value Stream Mapping.
When you say, “Trying to make change in Model I is difficult because you are dealing with their Espoused Theory”, I’m not sure that’s completely accurate. You are dealing with someone’s Model I theory-in-use. The difficulty in moving from Model I to Model II is that people are unaware of the gap between what they espouse and how they act. If you ask them, many will say that they believe they already use Model II (e.g. it’s their espoused-theory).
You go on to say “It is neither rewarding nor safe for a person to explore or actually change their mental models and decision rules, so there is a wide gap between their Espoused Theory and their Theory in Use”
I’m not sure of your data that it is not rewarding for someone to explore or change their mental models. Perhaps you are arguing that it is not rewarding if they find the topic potentially embarrassing or threatening? In which case, I’d agree that it’s unlikely to be immediately rewarding.
According to Argyris, a reason for the gap between someone’s espoused theory and their theory-in-use is that we are ‘blind’ to it. Argyris calls this ‘skilled incompetence’. This is why Argyris talks about the need for valid data, such as tape recordings, or imagined dialogues to help people believe there is a gap. Even once people get over their ‘blindness’ to operating with a Model I theory-in-use, the challenge is that they often don’t want to give up the Model I view of ‘being nice’ and ‘looking after others’ which Argyris argues are beliefs and behaviours we are socialised with as we grow up.
From my study of Argyris, it’s not accurate to say that Model II is the avoidance of “seeking to impose a view upon the situation”. Model II, an open to learning approach, is not the opposite of Model I ‘unilateral control’ position. This is similar to saying that active listening to another person is not just the suppression of the urge to talk. Model II uses the approach of ‘strong views, weakly held’. A Model II approach is one where positions are put out there to be disputed in an open way. The phrase you use, ‘open to exploration by others’ carries a potential meaning that it’s the others that need to learn. Model II advocates testing the ideas because we may also be wrong ourselves.
You argument is that by having someone participate in the data collection exercise they are more likely to believe the data that they find (satisfying Argyris’ first requirement of an intervention, ‘valid data’) and consequently respond in a Model II manner. I think your second quote from Seddon also supports this notion that people are more likely to believe data is valid if they are personally involved.
I’m not convinced that simply being involved in uncovering the data is enough to shift someone’s mindset. My data for this is that Argyris mentions that even when you tell people about Model I and Model II and illustrate their current level of ‘skilled incompetence’, they are not able to produce Model II behaviour without focussed practice (he argues it takes the same amount of time as someone learning to play a decent game of tennis).
Argyris’ criticism of TQM management projects was that even when they did result in improvements, there was often no time spent asking the question “how long did we know about the wastes?”. When he asked this question himself, he reports people saying “We’ve known for a long time, but you don’t want to go into that because it’s a whole can of worms”. In other words, even when programs have been successful, they haven’t uncovered or deal with the organisational defences (due to Model theories-in-use) that created the situation in the first place.
My understanding of the Vanguard work, is that they go through a process where they ask, “given this waste, what was the thinking behind it?”. The Vanguard Guide to Process Mapping and Analysis also has the following:
By asking ‘what are the management assumptions about the work, the people who do the work, and the way the work is done?’ you will identify the thinking that drives behaviour.
Stuart Corrigan’s (Vanguard Scotland) guides show an example where they suggest writing out the de-facto policy that would explain the waste and then identifying the thinking behind it.
I think that this exercise of directly looking at the thinking (based on data that the participants believe is valid) is the key to inviting people to test their current view of the world (mindset) and decide if they want to adopt a different set of beliefs (i.e. Double Loop Learning).
If you’re arguing that the reflection on the thinking is part of the Process Mapping process, then I’d say it’s not fair to compare it Value Stream Mapping which is much more of a tool than a process.
My view is that it’s the process of examining the thinking, that is illustrated by data that participants believe is valid, which leads to the difference, and not the tool itself. In this way, I could imagine the same reflective thinking process being used with VSM with similar effect. I’d be excited to understand if you have data (which I’d consider valid) illustrating that it’s less likely with the VSM tool.
In a Model II spirit, I’d welcome you challenging this view 🙂
@Benjamin
You say “If you’re arguing that the reflection on the thinking is part of the Process Mapping process, then I’d say it’s not fair to compare it Value Stream Mapping which is much more of a tool than a process.”
I think we’re working with different definitions of VSM here or at least different experiences with it. In my experience, VSM is not just a tool and it’s certainly not just a map – it’s a process similar to what’s being described here with process mapping. We’re quibbling over terms and I think drawing a line around “VSM is bad” but “process mapping is good” doesn’t move the discussion forward.
I’ve seen plenty of good thinking around VSM. People are involved in looking at the process and seeing the whole system and they believe the data — the things that you’re attributing to Seddon and process mapping also occur with Lean and VSM. Some of you reading this won’t believe that, but it’s true.
I’ve seen plenty of bad thinking around process mapping (which I could just as easily portray as a “tool”) but I’m not going to say process mapping automatically leads to bad behavior or closed thinking. Some people use VSM badly and some people use process mapping badly. Some people use either method to drive amazing insights and system redesign.
We should be pointing out bad thinking or bad management where ever it occurs, not making generalizations about which approaches/tools/processes/mindsets (whatever you want to call them) are good or bad.
It probably would have been more accurate to describe this as VSMs created in workshop settings vs ones created in the field.
Argyris reminds me of the model used in Crucial Conversations / Confrontations. I’m wondering how it relates to what we know about influence and approaches like Positive Deviance.
I’m not entirely sure about people always acting consistently with a Theory-In-Use. From what I understand, the factors that influence how we behave are not that conveniently simple.
Nice provocative title, David 😉
In my experience, the process of creating a VSM with any group, whether it’s managers or front-line workers causes lightbulbs to go on in people’s heads when they have to share and combine their own models of how their own area of responsibility fits into the bigger picture.
I’m sure if you then took that VSM and used it to beat on people who were responsible for dysfunctional parts of it, you’d find it was counter-productive. But using any tool as a bludgeon is counter-productive, right?
One thing I always think is important to point out is that the VSM can and will change, either as you tweak the system to improve it, or simply as a better understanding of the current system emerges. Treating it as a living model which is just used to generate insights for improvement, in my experience, means people don’t feel threatened by it.
Great post! Can you post more on the Process Mapping mechanics? Thanks!
Provocative title encouraged me to read the blog, however when I lead teams in VSM I do involve the workers and their peers in the development of the VSM by going to the workplace (gemba if you must) not with management locked in an office. Very few of the managers I have worked with blame their subordinates and/or the workers because they are big enough to understand they are responsible for the systems – further they want to encourage improvements not shoot the messengers. I often preface my leadership of VSM work by calling it an enhanced process flow map, on the basis it contains extra information and the data thereby collected is useful for task balancing and other improvements, if not immediately then in the near future. However you have started an interesting discussion – if this blog discourages people from VSM it will be a pity.If it encourages people to better pre-position VSM it will have been a success.If it encourages the use of simpler process flow maps in the absence of anything at all it will be a success.Looking forward to the next post!