Wednesday, July 31, 2019

What culture is right for you?

During 2005-2006 with the implementation of a pioneer-settler-town planner structure in a high tech startup in Old Street, London (a barren wasteland of technology at that time), one noted observation was that the different components of the structure appeared to have different cultures. A more recent set of observations (2014-2016), through examination of company’s doctrine has shown us that the principles used in companies are not the same and not even necessarily uniform within a company and this appears to impact the adaptability of the organisation.

To put it simply, if I take a list of universally useful principles derived from mapping (see figure 1) and then compare companies with a simple traffic light system (are you good at this or not) then differences exist (figure 2 and 3).

Figure 1 - Doctrine, a list of universally useful principles.

Figure 2 - Doctrine within a Web Giant

Figure 3 - Doctrine within a Banking Giant

Now, we cannot say which principles matter more and the doctrine list itself is constantly expanding. For example, as technology evolves then there is a constant co-evolution of practice and some of those practices are instead universally useful principles i.e. focus on user needs or use appropriate methods (i.e. XP vs Lean vs Six Sigma).  

It's worth unpacking this a bit more. Practice itself tends to be tied with a specific implementation of an underlying systems i.e. best architectural practice for compute as a product included disaster recovery tests and capacity planning. The practice itself evolves - it doesn't appear fully formed - and hence we have novel, emerging, good and best practice i.e. all four stages of any evolving capital (see figure 4).

Figure 4 - The stages of capital and the labels we use.

When technology evolves, a novel practice can co-evolve due to some characteristic change. In the shift from Compute as a Product (stage III, see figure 4 above) to Compute as a Utility (stage IV, see figure 4 above) then the characteristic that changed was MTTR (mean time to recovery). In the old world, when a server went bang it could take months to get a new machine. In the new world, it can take seconds.  This allowed for a new set of practices (which we labelled, overtime, as DevOps) and included design for failure, continuous deployment, distributed systems and chaos engines (random introduction of failure into the system). Such practices then themselves evolved and are now considered good (becoming best) architectural practice for compute as a utility.

This is best seen with a map, hence I've highlighted the process in figure 5. It's worth noting that a major form of inertia to a change of an underlying technology are the best practices that existed with the less evolved technology. It's not just pre-existing capital (i.e. sunk cost) that causes a resistance to change (such as shift from product to utility) but also the practices associated with the past way of doing things.

Figure 5 - A map of co-evolution with technology.

Sometimes, within the maelstrom of a changing technology there emerges a set of practices that transcend the context and are universally useful i.e. focus on user needs or think small i.e. know the details.  Sometimes these practices are hidden in more context specific practices i.e thinking small hidden in both micro services and the use of cell based structures. These practices are not dependent upon the underlying technology and the specific context.  These practices we call universally useful principles and doctrine is simply a collection of universally useful principles. The appearance of such principles is often by happenstance or as a process of people rethinking how we do stuff.  The principles do have a contextual side to them i.e. my user needs are not necessarily the same as your user needs but overall a focus on user needs is seen as universally beneficial.

However, what matters in competition is also your competition. Being hopeless at these principles doesn't matter as long as your competitors are the same. If we all ignore user needs then no-one gains any advantage. It only becomes a problem when a different style of company (such as the Web Giant) enters our industry (i.e. Banking). Finally, within a single company these principles aren't even uniform. Some parts may well have a different style of leadership and a different emphasis.

Overall, these statements should not be surprising - "cultures and related principles are different between organisations and within organisations and this may have an impact on performance".

Unfortunately, this leads us to a question of “What culture is right for you”. The problem with that question is in its unpacking.

Firstly, we have to define what culture is in order to distinguish between what maybe right or maybe wrong? In 1952, the American anthropologists, Kroeber and Kluckhohn, critically reviewed concepts and definitions of culture in the academic world and compiled a list of 164 different definitions. Any review of popular business press shows the term is equally, if not more so nebulous today. Despite this, we assign almost mystical properties and powers to “culture” with memes such as “culture eats strategy for breakfast”.

The subject matter itself is complicated in terms of many different components that influence it from learned to inherited behaviours, from both social to personal constructs, from layers to manifestations, from determinism to relativism, from the semantic to symbolic history. It is also complex in that many of its attributes appear emergent as with many evolving processes. The field is also littered in business press with many inadequate concepts that do no hold up well to scrutiny such as the idea that culture is homogenous (as in one thing), that it is evenly distributed among members of any social group (including organisations), that individuals possess just a single culture or that culture is simply custom or that somehow culture is fixed rather than transient and timeless.

At best, we can say that culture contains many things from experience, learned behaviour, knowledge, meanings, relationships, hierarchies (i.e. power structures), capabilities (i.e. skills), values, possessions, aspects of belief (i.e. ethical position relative to others), principles and attitudes. It has many layers from the individual to the group to the organisation to the nation. It has many perspectives which may vary with context.

In terms of the question, the problems with it can be stated as follows :-

1) with the word “culture” and its definition.
2) with the words “what is right” as this requires some measure of what is right and acceptance that there may be many rights and many wrongs both of which may change with context.
3) with the words “for you” as there are many perspectives. There is also the derived implication that culture can be directed because there is no point in knowing what's right if you can't get there. The issue here, is that many parts may only be ever emergent i.e. we might be able to nudge but can't force a direction.

This mess of a subject however does not prevent a thriving and lucrative industry of self-help guides, performance value matrices, eight or ten step processes and declarations of what makes a successful culture - short term wins, empowering employees, establishing a sense of urgency, communicating the vision, from good to great matrices etc. Most disturbingly, many of the characteristics described as associated with success often through positive case examples can equally be found in those that failed. This also leads to a notorious problem of outcome bias where we adopt some approach simply because others we see as successful have used that approach and finally confirmation bias where the evidence that such an approach works is that others are now adopting it. This seems to lead to an endless barrage of memes and the desire to get "big names" to bless some approach. 

From my personal experience, the net result of an extensive reading of business press left me with one of two paths, either join some cult of belief around the latest management guru or throw my hands up in horror and retire to the marsh I inhabit. Fortunately, anthropology came to the rescue. The single distinguishing difference between anthropology and the business press when it comes to the issue of culture is that anthropology is observation led and focuses to minute detail on observation whereas the popular business press seems to start with a hypothesis and find case examples to fit it.

This question of observation provides me with a route into the field, because maps are all about observation. The very start of the strategy cycle (figure 6) is the first O of OODA - observation, from which contexts can be learned and patterns discovered then reapplied.

Figure 6 - The Strategy Cycle.

It's not just activities and practice that evolve. From figure 4 above, we could map knowledge as a form of capital. As a primitive example, I've added a quick map of population genetics using the knowledge axis. Now, you might disagree with how evolved the components are, you might add additional components that are missing but that's the point of maps. Our evolving higher order systems are built upon evolving lower order components whether activities, practices, data or knowledge.

Figure 7 - Population Genetics as Knowledge.

In the same way, we have to accept that culture has many components rather than being one thing. It is a construct built on constructs. The question is, can those components be mapped or can mapping help? We already know that mapping has proved useful in terms of understanding a competitive landscape, investigation of practices, determining what appear to be commonly useful principles (i.e. doctrine) and examination of other forms of capital (i.e. knowledge or data). Maps inherently have perspective and many layers can be used from the individual to the nation state. Maps are designed around the concepts of evolution and for this reason maps can assist with both the complicated and the complex (emergent behaviours).  Maps are a context aware mechanism of communication and learning for both past actions and potential future behaviours. 

At first glance, these concepts of context and of universal patterns appear to chime well with the ideas of culture. We already have some fair advanced thinking on landscape, on doctrine and on inertia. We can even expose attitude through organisational structures such as Pioneers - Settler - Town Planner.  But there are limitation with maps. The technique is imperfect (all maps, in order to be useful, have to be imperfect representations). We also know that maps themselves can change behaviour, from the tackling of bias to inertia. Maps have the potential to be a visible artefact of culture and as much a part of the landscape as a means of visualising it.  The most severe limitation is that maps depend upon evolution of capital.

However, ideas, the meanings between them and even our values and ethical constructs evolve. In Edgar Schein’s Organisational Culture : A dynamic model he directly refers to the changing nature of value from espoused values to assumptions to unconscious responses.

In mapping terms, a similar evolutionary scale can be acquired by combining the labels from different forms of capital (across practices, data and knowledge). Taking figure 4 above, I've highlighted the following labels (figure 8) to provide a scale of  :-

Stage  I: Concept (and observation) - an espoused value.
Stage  II : Emerging - the evolution of an espoused value as an emerging theme.
Stage III : Converging - the conflict between emerging themes to form an assumption or an array of assumptions
Stage IV : Accepted - the unconscious response when the once espoused value has become an accepted norm and taken for granted.

Figure 8 - Labels for ethical values.
As with any form of capital, we can simply map this and combine it with other forms of capital. Using the ubiquitous cup of tea example, we can add to the discussion not only the physical components, any practices, any knowledge, any flow of capital (i.e. finance) but also considerations of value and ethics such as fair trading or green energy (see figure 9). In this figure I've changed the x-axis (in red) to the labels used for discussing ethical capital.

Figure 9 - Expanding the cup of tea.

Since the labels on the x-axis are simply notations for stage I to IV of capital, I have provided the alternative labels in the figure above but normally, I just use the labels for activities (from genesis to utility) as in figure 10 whether I'm talking about activities, practice, data, knowledge or even ethical values. This provides a good enough representation for me.

Figure 10 - A normal map including ethical values.

Since mapping itself is derived from the strategy cycle, it is inherent in maps that the existing state is derived from looping around the cycle i.e. past action informs the present. This would mean that culture (as with other forms of capital) in an organisation would not only be a set of evolving components that interact but also a consequence of past actions (i.e. loops). The choices of today being limited by past contexts and decisions. This means that culture cannot be measured against some standard because it is dependant upon the landscape, varies with history and is inherently emergent in the same way that strategy is. 

However, this doesn’t mean patterns don’t exist and hence the same climatic patterns (rules of the game), context specific patterns (methods of manipulating certain contexts) and universally useful principles (i.e. doctrine) can exist. 

Finally, mapping does distinguish between the needs (it describe a chain of needs in a landscape) versus the want (i.e. often how we wish to influence or manipulate the landscape). This difference of wants and needs is also present in culture i.e. the a set of generic wants of the poor and the middle class might be identical but the needs can be very different. All has a cultural bearing when it comes to describing a landscape.

Though all maps are an imperfect representation of a space, they have uses in communication, learning and pattern recognition. Some of the components of culture (whether principles or attitudes or ethical values) can be mapped with the assumption of values evolving as described above. Some of the components of today's culture are a consequence of past choices in past landscape. In the same way that we have technical debt to past choices and inertia to change from past success, we should also see cultural debt and inertia to change from ethical capital.

Of course, I've specifically focused on activities (one type of capital) with a particular reference to technology in this discussion whereas co-evolution can occur with many forms of capital i.e. co-evolution of practices (and hence the appearance of principles) can occur due to evolving knowledge or data or even practices themselves (in a delightful recursive twist).

This, however, is a dangerous line of inquiry. As we already know, competitive markets can be easily mapped and manipulated through gameplay from a local industry to nation state. The problem is not can we map and manipulate the culture of a local organisation to create some favourable change as we can simply do this by giving someone the list of doctrine and ask them to examine the company. The problem is can we manipulate the ethical values of a wider system such as a nation state once we have mapped it?

For over a decade, I've been hesitant to explain more on this topic. Times have changed though. Mapping has spread enough that I should open the door a bit more.