Systemic coaching


Systemic coaching

I see an increasing number of references to, and sales pitches for, “systemic coaching” and in this short article, I’d like to reflect on what a coaching approach would need to be like to justify the claim to being “systemic”.

Before starting on that, though, a couple of qualifiers. First, I’m writing here as a systems specialist not a coaching specialist. In fact I don’t see myself as a coach at all, I certainly don’t have any coaching qualifications and I don’t coach as a job. Second, all that said, I’m not in any way going all Animal Farm and assuming “systemic good, non-systemic bad”, but I do have a, perhaps old fashioned, view that if you advertise something you are selling as having some particular feature, it should include at least a bit of that in the package. I’m not setting up a systems way of thinking as being intrinsically better, but it is intrinsically different. If you are advertising “systemic coaching” then it should be recognisably very different to conventional coaching and built on some different principles. It also follows that if you want a different approach to coaching, then thinking about it systemically should be a sure bet to come up with that.

So, to systems and coaching and more specifically to what makes any approach systemic in nature. Systems approaches vary a lot, but there are some very consistent threads – recognisable and consistent ways of thinking that weave their way through most systems approaches and which are fundamental to thinking systemically. Some of these seem particularly relevant to a coaching context.


Systems approaches are about seeing wholes rather than parts. Often this is glossed over in quite an “arm waving” way, but it’s actually a tight and well defined pattern of thinking that runs counter to how many people have been trained to think as a default. Faced with something that we want to understand, the common and analytic approach is look at the parts of the thing we want to understand, and build up our understanding of the whole thing from the understanding of the parts. Systems doesn’t work like that. Instead we take the thing we want to understand and, rather than asking what its parts are, we ask instead what it is a part of. Then we build our understanding of the thing of interest from first understanding its context, not its parts.

This systemic thinking technique got carried into therapy and became the basis of family therapy, where we seek to understand an individual and their behaviour primarily through understanding the family as a whole, and then the individual from their role within that.

Take that fundamental thinking principle and transpose it into a coaching context and we would focus first on the client’s organisational context and understand them from that perspective rather than focusing primarily on the individual’s needs behaviours or goals. We’d seek to understand the individual’s behaviours as an aspect of the context or organisation they are in. In practical terms, then, we’d want to understand the ways in which organisations act to change individuals and what we can do about that, rather than focusing mainly on their internal state.


This is one of the hardest aspects of systems to hold onto. In systems the object of our focus is on connections more than on the things the connections are between. This isn’t too difficult as a concept, but in practice, it’s really hard to stay focused on the relationship rather than slipping into a “things based” view of the world. No least because so much conventional thinking in this area is “things based”. Occupational psychology, for example, relies very heavily on psychometrics which provide different ways of “typing” individuals that neatly classifies them as a certain type of person, a peg with particular characteristics of round or squareness.

The connections view is very different and using that we’d characterise individuals not by their inherent “isness” but by the nature of their relating in the particular context they are in. There is a really substantial body of evidence both from experimental psychology and practice to support this approach. One well known example is the Stanford “prisoners and guards” experiment where students were given one of two roles and famously, the experiment had to be stopped as the “prisoners” were turning feral and the “guards” increasingly sadistic. Irrespective of any inherent fixed characteristics that individuals may or may not possess, in practice, their (our) values and behaviour are demonstrably incredibly malleable and formed by relationships at a surprising rate. Outside of clinical experiments, many people who have worked with management simulations will know just how powerfully and how fast these can norm behaviour. Just as one example, in Barry Oshry’s Power and Systems lab, the structure of relationship consistently determines how individuals will act, feel and how they will see the world around them and this has been demonstrated in thousands of cases. Whatever their inherent characteristics, people behaved, thought and felt in very consistent and predictable ways. So relationship overrides characteristic.

So how would that translate into a coaching context? Rather than seeing the individual as having inherent characteristics, we would instead focus our attention on the nature of their relationships. To do that well, we’d need some tools for modelling relationships which offer reliable ways to understand how to change them to advantage. A relatively few such tools do exist, Transactional Analysis, for example, which is used extensively in coaching in some parts of the world.


Systems approaches focus heavily on dynamics and on understanding how those unfold through time. This relates very closely to the previous point about focusing on relationships rather than the things the relationship is between. As the great systems thinker Gregory Bateson pointed out:

 “The horse didn’t evolve, the field grass didn’t evolve. It is the relationship that evolved. The horse and the tundra with grassy plains are interlocked. It’s an evolution in which the grass needs the horse as much as the horse needs the grass. “

So change is driven by the dynamics of relationships and, in a coaching context, I think this has several implications. It means we need to be able to model or at least understand those dynamics and where they are taking us. As the individual client co-evolves with their organisational context, is that taking them somewhere they want to go? If the dynamics of the relationship are trapping them and preventing progress, what is the nature of that trap? Sometimes you have to understand the nature of the trap in order to spring it and allow change.

Of course, this is also reflexive. For us as practitioners the same applies in our relationship to our clients whether that’s as a coach or a consultant: is this relationship moving them forward, or is it holding them in place? At what point do palliative measures to alleviate a difficult situation become complicity in maintaining the stasis, or worse parasitic – maintaining the problem for our own benefit? To be genuinely systemic then coaching would need tools to model and help manage the dynamics of change and some of the standard models used in organisational change – the adoptions bell curve and bereavement curve – are inappropriate (because they were developed for other uses and in very different contexts) and inadequate for doing this. The tools that we know work for modelling dynamics of relationships tend to be a bit heavyweight for a coaching context because they’ve mostly been built for dealing with organisational problems, but some recent research by a PhD student at Manchester Business School holds promise in providing a way to pinpoint and track how individuals’ values and thinking affects and is affected by their organisational context.


One of the less understood characteristics of systems approaches is that they are based on building models. As systems practitioners we build models (it does attract a lot of geeks), then try “what if” changes in the models before experimenting on reality. This is because we’re generally dealing with things that are so big and complex and with such long lags in effect that this is the only way to deal with them effectively. Obviously, it’s not just possible but perfectly normal to intervene with an individual without using any sort of model. The relevance of modelling for coaching is, I think, that it brings an element of transparency for the coach and transferability for the client. If our anxiety as coaches (well my anxiety as a consultant) is to know that what I’m providing is as good as it reasonably could be, then having a model of what I’m doing makes the comparison against alternative approaches at least feasible. If our objective is to help clients grow and flourish, then that implies them becoming less dependent on us and models become an easy way to transfer learning. Without them, we’re reduced to dark arts, and an inability to explain why we suggested X rather than Y or why Z did or didn’t work.

In conventional coaching this “modelling deficit” is partly covered by supervision and the role of a supervisor working with for example Shoett and Hawkins Seven eyed model allows for reflection from a more systemic perspective. However, we believe that coaching contracts designed to embrace all seven eyes are scarce.


One of the implications of systemic coaching focusing on the context and on the dynamics of relationships of the parts in the system is that this gives you many more places to intervene. Where, in conventional coaching, intervention is aimed at the individual, a systemic approach could involve intervening on the individual, the organisation, the relationship between the two, or any combination of those three. Many more options, and that increase in options has to increase the chances of success. Reg Revans – the father of Action Learning – found that interventions to change the organisation were on average five times as effective as interventions to change the individual, but of course individual and context like Bateson’s horse and grassland are inextricably intertwined, so having the options to intervene in both maximises your chances for being effective. For fans of Ken Wilber, a genuinely systemic approach to coaching would go a long way to fulfilling the needs of his AQAL model, and offer both intervention approaches in all four of his quadrants (internal / external & individual / collective) as well as providing a way to model how each quadrant was affecting the others.

Another implication is that to build an approach to coaching that is genuinely systemic will require developing some new approaches, but perhaps not that many. There are well tried and trusted approaches in the systems practitioner’s toolbox that cover the contextual side of this relationship, there are some (but probably not enough) approaches in the coaching toolkit for modelling relationships and there are at least a couple of approaches for modelling dynamics – some a bit clunky, some reasonably slick. What is required though is working out how these might fit together and where the shortfalls might be. So it is one thing to be able (as we are) to diagnose an organisational situation systemically and flag up the underlying systemic issues, the probable symptoms and even the likely effect on individuals. But it’s quite another matter to be reasonably sure about whether intervening to change the individual, the organisation or the relationship between the two is preferable and how that might play out through time and what the effects on each would be.

On that I feel we’re just beginning. We are involved in a multi-disciplinary initiative (the systemic coaching development project or SCDP) involving exec coaches and systems practitioners to develop just such an approach, if you’d be interested, we’d love to hear from you at .