Introduction to Soft Systems Methodology: A Holistic CI Approach
Continuous Improvement isn’t just Lean, Agile and Six-Sigma. There are other disciplines that fit under the CI “brolly” for instance Practical Process Improvement, Organisational Development and Systems Thinking.
Systems Thinking grew out of cybernetics. There are two approaches discussed in Systems Thinking, the Hard Systems Method (HSM) and Soft Systems Methodology (SSM), Systems Thinkers may also talk about the Viable Systems Model (VSM).
HSM assumes a situation is well defined and has a single optimum solution, it is a scientific approach. In the discipline of Systems Thinking this was all that was discussed from its inception until Professor Peter Checkland (from the University of Lancaster) realised that HSM thinking was not enough, for many systems to function properly they had to take account of the humanity that the system has to deal with. This is where the difficulty lies, people aren’t as predictable as machines. To deal with humans you had to understand how the system worked before deciding how to change it. Checkland came up with the Soft Systems Methodology, which is the approach that most people think of when referring to Systems Thinking.
You will see this approach look fairly similar to plan-do-study-act and the Lean approach of:
1. Study the Current State and define the problem statement
2. Define the ideal state
3. Analyse the gap between current state and ideal state
4. Map the future state
5. Create implementation plan and
6. Implement and measure
7. Improve the improvement process
Checkland split the SSM into two parts, real world and systems world. Initially the person making the study would try to understand and express, or model, the current real-world situation. Different models might be: Rich pictures, Systems Maps, Influence Diagrams, Multiple-cause diagrams, each of these must be from a specific point of view, and this, therefore assumes that the person creating it has been immersed in the situation, in Lean terms “has gone to Gemba”. In this case going to Gemba is not just going to where the work is done but seeing all the effects of the system and in doing so talking to all the stakeholders to hear their views on the system.
The whole point of these models is to be able to view the system systemically, that is holistically. What Systems Thinking tries to do is to try to see the consequences, positive or negative, of any change before it occurs. Therefore one of the things that Systems Thinking tries not to do is reductionism, i.e. tackling the improvement of the system by breaking it into bits, solving the problems discretely, then putting them back together and expecting it to work.
So what is a system?
A system is a bunch of stuff (things, actions, people) that when put together in a certain way equal more than the sum of the parts, in Lean language, adds value.
Boundaries
When I was a kid I would sometimes write my address as Darren Clyde, 30 Belhill Road*, Belfast, Antrim, Northern Ireland, United Kingdom, Europe, The Earth, The Milky Way, the Universe, whilst technically correct much of it was not of interest to the postal system. Each of the elements of this address are different boundaries, with “the Universe” being the ultimate boundary (not taking into account the multiverse hypothesis), while something happening in Belfast will affect the universe as a whole the size of that effect is so miniscule as to not be worth considering. This is the same when studying any system, while it is important to see the system holistically, to be effective sensible boundaries need to be set. When my Dad was renovating the house in the 1980s his system for purchasing building supplies extended to the city of Belfast and counties Antrim and Down. If he had been doing it today, with eBay and other online purchasing options, the boundary of the system could have been the whole of the British Isles, and perhaps, for certain items, The Earth.
Boundaries to systems aren’t just geographical, they can be time, money, people (quantity, competence), purpose (viewpoint), risk, etc.
So we can now set the boundary for our system of interest and study the particular problem and its environment. Let’s say our problem seems to be my local shop is losing customers. Our boundary is the people living and working in the neighbourhood. I can currently afford to pay one person to work in the shop, I have a few thousand pounds saved up, and the purpose of the shop is to provide small amounts of “necessaries” to our customers.
System root definitions:
We have modelled the current situation we now need to understand the system’s root definitions. In Systems Thinking the value, or purpose, of a system is subjective. Take a car:
- The designer views it as a machine to take a given number of people and/or cargo, from a point A to Point B in a certain time in certain conditions to a certain level of comfort and safety.
- The car company finance director might think of it as a method to increase revenue and possibly see it as way to drain cash reserves from the company
- To an Uber driver it might be a way to make money
- To an private owner it might be a convenience that can be called upon to make their lives easier, or a thing that raises their standing in the particular community they live in (perhaps the vehicle is a zero carbon car and the person is a member of Greenpeace).
These are relatively positive views, but there are negative purposes to the system:
- A way to pump as much CO2 into the atmosphere as possible
- A way to increase traffic congestion
- A way to build an argument for more roads
- A way to increase road deaths
Granted these negative purposes are probably not the design aims of cars, but it illustrates that defining the system and its purpose is important. Early in systems thinking it was taken that the purpose of a system was what it was designed to do, leading to a blindness to negative effects. In the turn of the century a noted Systems Thinker, Stafford Beer, challenged this thinking by saying the “Purpose of a system is what it does” or POSWID. This means when studying a particular system the practitioner must look at the bad stuff as well as the good, see the risks as well as the opportunities. David Smyth (a researcher who worked with Peter Checkland) provided an abbreviation to frame system root definitions, CATWOE: Customer(s), Actors, Transformation, Worldview, Owner, Environment.
For the car example:
- Cusromer — The driver
- Actors- the designers, builders, and salespeople
- Transformation — Driver moving from point A to Point B
- Worldview — that it is acceptable to drive that type of car that distance
- Owner — the person who has paid for the car last
- Environment — On the road in all weathers except extreme snow, total ice coverage and flood.
As discussed, point of view is also important, so consider:
- Customer — the oil companies
- Actors — the motor designers, builders and salespeople, and the motor manufacturer executives
- Transformation — consumption of hydrocarbons bought from the oil company
- Worldview — that transforming hydrocarbons into CO2 (and other damaging by products) is acceptable
- Owner — the person who paid for the vehicle last
- Environment — On roads with access to the company’s filling stations.
All though many of the words are the same it paints a very different picture, now we are able to have a different discussion on the purpose of the system and can now consider different options to improve it.
The idea is to make as many different purpose definitions as possible to pull out all the positive and negative views of the system (do this for all the stakeholders you can think of, and bear in mind that there may be more than one per stakeholder).
Let’s go back to my shop:
From the different definitions we can now have a different take on the problem. The customer is looking, for goods that they have forgotten as well as cheaper goods. My worldview is a local shop is a valid and valuable service. There is clearly a disconnect. The problem now has more definition because of the analysis of the system root definition. We can now model solution approaches to see if we can either shift the Transformation or bridge the disconnect.
These models are likely to initially be the same format used to model the current state, e.g. rich picture; try to conceptualise different solutions, 7 is a good number, up to six attempts to improve the system are likely to be variations on a theme, the 7th, because we would have run out of variations would, therefore, need to be a much more radical solution. For our shop the first six might be about reducing the costs, the 7th might be an online option that guarantees people have a shorter delivery time than the big suppliers on selected items list based upon analysis of historical purchases and volume of requests for specific items.
Step back into the real world
We can now test our ideas against the real world, it could be review them against our current state models, it could be compare to the actual work going on, it could be explaining them to the stakeholders and getting their views, often it will be all of the above.
From there we can assess the ideas against the problem situation, how do the solutions address the problem? What unintended consequences can we see, weighted criteria selection chart might be useful here.
Once we have chosen the solution, we can try to improve it by taking good ideas from the other proposals and adding them in to the chosen one. The next step is to list the actions that need to be done for the new solution to be embedded, now execute. In the local shop this might be a series of projects e.g a project to set-up a website, or a project to reduce overheads or a project to get items to our customers in an acceptable timeframe.
Continual Improvement
It is a concept in Systems Thinking that systems are always changing, so need to be constantly monitored, day-to-day this is done with performance measures, however measures are only one model of the system, the system needs to be properly modelled regularly to stay relevant, this means as soon as the SSM cycle has been completed the Systems Thinker, like the CI practitioner, is already looking for the next problem to solve.
So, what does this mean? Typically, the classic CI approaches like Kaizen and Six-Sigma are great at solving discrete problems but they might not be the right problem; you might be “polishing the cannonball”. Systems Thinking is good at realigning your organisation and deciding where to aim your Kaizen or DMAIC improvement projects.
Darren Clyde has spent 15 years working with organisations and teams to reduce cost and effort of doing business. He has taught staff of all levels in the use of Systems Thinking and the Soft Systems Methodology and how they mesh with the more usual CI approaches.
*I made up this address to protect the innocent.
