Assessing Culture in Complex Systems

Understanding culture: 

The problem with surveys: from evaluation to description

One of the key shifts required in organisations is how we view culture and culture transformation.  If one studies recent business literature, it soon becomes apparent that executives have realised the importance of culture.  “Culture eats strategy for breakfast” has become one of the most-used phrases around.  Another often used phrase is “soft stuff has become the hard stuff” meaning that executives hopefully realise that culture can no longer be seen as soft, unimportant and something that can be delegated to OD or HR.  Even with this new realisation of importance, and all the various culture models and culture change tools around, sustainable cultural transformation still seems to be elusive in most organisations.

I believe that a big part of the problem lies in the way we think about culture.  In most of the organisations I work with, culture is still seen as something that can be accurately diagnosed and designed.  There is a belief that the leadership team, or a group of expert consultants can design and implement an objectively ideal culture.   The issue with this is that culture is an emergent property of a complex system.  It emerges and evolves over time from a myriad of interactions between people, their environment and the structural aspects of the organisation e.g. processes, policies etc.  It finds expression (and is shaped by) the ongoing conversation, the stories and observations shared in casual conversation, the symbols displayed and the rituals engaged in, sometimes subconsciously.  It often cannot be fully articulated, but it known by everyone who is a part of the system.

Because the culture is ever emerging and evolving, any attempt to accurately “diagnose” it is flawed.  Any culture audit result is only ever a partial point in time description, and the very act of diagnosing changes the system.   The way we go about the diagnosis is also problematic.  From a complex systems perspective,  typical culture audit instruments has several short comings:

1.     They are evaluative, not descriptive

A typical culture audit instrument asks people to evaluate the various aspects of the culture.  For example, staff might be asked to evaluate their leaders in terms of whether or not they are appropriate role models to follow, or whether they live the organisation’s values.  The problem with this is that an evaluative process is always open to gaming.  Because it is very apparent what the “right answer” is, people could either gift you’re the answer they think you want, or game the process.  Others simply go into an “automatic pilot” mode and simply got down the middle without thinking about the answers.  Also, being asked to evaluate vs describe something immediately puts someone in a very different state of mind.  Often we become more critical in an attempt to be more rational, often looking for flaws or reasons to complain.   Evaluations and opinions are useful, but they are highly subjective.  It is much more useful to have a rich description of the experiences or observations that inform these opinions, because those are specific enough to act on.

2.     Questions lack context

In most of the surveys I’ve completed, I’ve wanted to answer: It depends … Sometimes I feel my leader is a good role model, sometimes he isn’t.  Sometimes I feel he lives the values, sometimes he doesn’t.  This leads to a “middle of the road” answer that has very little utility. 

3.     They are based on a specific model or hypothesis

Most culture auditing tools are based on someone’s model of what culture is and what the ideal looks like.  Anything that therefore falls outside of that framework will not be represented in the output.  I.e. we will only get the insights we thought to ask for.  Recently a client decided to deploy a culture audit tool developed in the US.  Seeing as the overt diversity and transformation issues we have here in SA is not as salient in the US, the instrument didn’t touch on those aspects at all.  Subsequent narrative scans revealed that transformation and diversity issues were some of the most urgent to address in this organisaition’s culture.  The auditing instrument completely missed it because it didn’t fit the underlying framework.

Because culture emerges from our interactions and the stories we tell about them, one way to get below the surface (to take the iceberg metaphor) and tap into this ongoing conversation is by continually gathering micro narratives or observations from across the organisation.  The key is to ask people to describe the system, not evaluate it.  

Once we have a richly described cultural landscape, we are able to evolve the system forward based on patterns across multiple observations.  Some of these patterns may lead to large-scale interventions that span the organisation e.g. changing the approach to performance management.  However the data also allows us to evolve forwards based on small local actions aligned to a coherent direction.  We do this by enabling decision-makers across all levels of the organisation to access the stories and observations from their area, and then to ask a simple question: what to I need to change or do differently to create fewer observations like these, and more like those? 

(This map was created with  Sensemaker )

(This map was created with Sensemaker)

Figure 1: Each Business Unit has their own unique landscape. (This map was created with   Sensemaker  )

Figure 1: Each Business Unit has their own unique landscape. (This map was created with Sensemaker)


This is a question that someone on a board level understands as well as someone on the shop floor.  It allows the entire company to evolve the culture into a coherent direction, but with local autonomy in terms of how they do it.  It’s also an ungameable process, the only way to change people’s observations or experience is to actually change something in the environment or in the way they interact.  

So we don’t need complicated (linear) change models, we map the narrative landscape through observations, give decision-makers across all levels of the organisation access to the resulting patterns and underlying stories and ask our simple question.

Notice that we don’t design an ideal culture for a potential future state that might never come about.  We evolve forwards from where we are today and learn and adapt as we go.

This is a big mindset shift – for leaders and practitioners alike that challenges our need for control and the plans and linear processes we are so comfortable with.

This guest blog was written by our recent guest, Sonja Blignaut, who is a narrative practitioner, change catalyst, and founder of a niche consultancy called More Beyond. You can listen to her interview on the HumanCurrent podcast here and don't forget to subscribe in iTunes.