Monthly Archives: December 2018

Investing with Cynefin: Disorder

Disorder is the fifth quadrant in the Cynefin model. Disorder is where there is no clarity about which of the other domains should apply.

“Here, multiple perspectives jostle for prominence, factional leaders argue with one another, and cacophony rules”.


The solution is “obvious”, bring the leaders together to perform a “complicated” ritual that reveals the “complex” nature of the problem and hope that the personalities involved do not turn the conversation into something “chaotic”.

A technique like four corners contextualisation can be used to facilitate a conversation between decision makers. The discussion will normally reveal that the problem is one of granularity. Decision making is a classic example of something that is often in the domain of disorder. However, when we break it down into a lower level of granularity, we discover that it falls in the other four domains.

  • The output, an ordered list is obvious.
  • The process such as cost of delay is complicated.
  • The interaction between the investment options, the available resources and the participants are complex.
  • The behaviour of warring factions is chaotic.

By moving to a lower level of granulation, the domains become apparent.

Investments that remain in the disorder quadrant indicate a dysfunctional decision making group. Often a hippo will force these items into one of the other domains. Disorder is often a symptom of a group stuck in “storming” that is not coming together to communicate.

Investing with Cynefin: Complicated

The complicated domain is the realm of strategic superiority. Organisations know with certainty how customers will behave, however it is not common knowledge. This is where organisations should focus their strategic investment.


The complicated domain is where organisations should be making larger investments as this is where they have competitive advantage. The only place where organisations should be making larger investments is where they are forced to due to regulatory dictate (Obvious) or to resist the vertiginous draw of the cliff (The Cliff). The complicated domain is where constrained resources should normally be deployed.

The complex and complicated domains are not binary in nature but rather “linear” with 0% certainty at one extreme and 100% at the other. As such, the investments are best managed using the Kelly criterion. The nature of the experiments change. Whereas in the complex domain, the experiments relate to understanding needs, in the complicated domain, the eperiments relate to the scope of the needs.

For organisations, the danger with the complicated domain is that too many investments are made in it. Either because they are classified incorrectly due to perverse cultural incentives or because the organisation is utterly risk averse. One is reminded of the risk averse anthem “No one ever got sacked for investing in the complicated domain. buying IBM.”. The real message being that perhaps some people should have been sacked for failing to think for themselves.

In summary, investing in the complicated domain is the easy option. Therefore the investment decision process should make it difficult to do so.

Investing with Cynefin: The cliff

In the Cynefin framework there is a cliff between the Obvious and Chaotic domains. Systems fall down this from the ordered domain down into the Chaotic domain. This is eloquently summarised by Mark Twain…

It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.


Retail businesses on the high street are currently thrashing around at the bottom of this cliff. After years of pulling levers to control shoppers and extract profits, the bricks and mortar retailers now resemble Dr Who randomly playing with the controls of the Tardis as it bounces around reality. The reason is that the internet has changed the fitness landscape of business and new apex predators have emerged that are disrupting existing players.

When organisations find themselves inside the OODA loop of an Apex Predator in an unfamiliar fitness landscape, they have to act, sense and respond. As a result, investments in the chaotic domain may result in disrupting the investment process in all of the other domains. Although investments in the chaotic domain should ideally be small, there are occasions when all the resources of the organisation will need to be focused on them.

Ideally organisations should avoid the cliff all together, and if necessary they will need to invest all of their resources to keep themselves away from it. Organisations can use metrics to detect the presence of the cliff. In particular they can use churn metrics. An organisation would typically have three sets of customer metrics…. Number of customers, customer activity and customer revenue. The trend is more important than the actual value for investment decision making.


In the example above the metrics may initially appear healthy. Every week sees a 10% increase (Green). However looking at the churn number indicates a problem occurred in week 4. At this point, the organisation needs to act as it heads towards the cliff / lands at the bottom of it. The action required is normally to gather information. Why has churn just jumped up? Which customer need is not being met, or possibly which customer need is being better met by a competitor’s product or service? The research will either involve data analysis or user experience research (to understand user needs/jobs to be done).

Many organisations wallowing at the bottom of the cliff have no data analysis or user experience research capability. For these organisations, the ‘act’ is simple, to acquire these capabilities.

Modern organisations wishing to avoid the cliff need to invest in data analysis and user experience research capabilities before they find themselves being disrupted by another organisation’s OODA loop.

Investing with Cynefin: Complex

For investments, the complex domain may be defined by the strategy used to manage it “Multiple hypotheses tested using safe to fail experiments”. The complex domain by definition contains uncertainty (risk) and is about acquiring knowledge. Investments the complex domain should be the smallest safe to fail experiment that test a hypothesis. Organisations should not be making investments that are large or that are not safe to fail.


Investments involving technology have three categories of riskĀ¹ :

  • Delivery risk.
  • Business Case Risk.
  • Risk of damaging the existing business model.

The Kelly criterion can be used to help understand the maximum size of investment in the Complex domain. Fundamentally investments in the complex domain are about reducing risk (uncertainty) by acquiring knowledge (certainty). The first two categories of risk are about failing to achieve an upside, they are not about protecting against a down side. Managing the risk of failing to achieve an upside is done by ensuring that individual investments are not so large that they damage the portfolio. Hence the Kelly criterion can be used. Whilst Kelly can assist with the first two categories of risk, alternate strategies are needed to protect against a down side.

Managing the risk of damage to the existing business model.

Protecting against down side is about managing the risk of unintended consequences. If a consequence is known, then a specific option should be created to ensure that the investment is safe to fail. Therefore ensuring investments are safe to fail, requires the investor to have options to detect problems and return the system to safety. This means the following:

  1. Effective monitoring is required to detect unintended consequences.
  2. Options to return the system to safety.
  3. Failure containment.

Failure to monitor for unintended consequences is an abdication of responsibility and normally indicates a risk averse culture dominated by Hippos. The Hippos either accept of ignore risks that might occur.

Time is the key element of options. If the time the system can survive is less than the time it takes to return the system to safety, more options are required before the investment should be considered safe to fail.

Finally, one of the key differentiators between contemporary organisations like and Google, Facebook and Netflix and traditional organisations is that contemporary organisations manage risk rather than ignore it. They create failure containment.Contemporary organisations roll out investments to customers gradually. They test hypotheses to ensure that not only does the investment work, but also the customers have the anticipated, or at least a beneficial, behaviour change. The idea that you would roll out an investment to 100% of your customer all at once is the equivalent of putting all of your money on “red” at the roulette table… You are not investing, you are gambling. You might get away with it a few times but eventually you will do a “Knight Capital”.

The Complex domain is fundamentally about risk management. As such, it is the domain where Real Options are most effective.

1-Original article by Steve Freeman and Chris Matts. Published in Agile Times, 2005

Investing with Cynefin: Obvious

The counter-intuitive aspect of investments in the obvious domain is that organisations should seek to minimise investment in this domain. As Dave Snowden once pointed out, organisations have no competitive advantage in the obvious domain. Investments in the obvious domain are often as a result of constraints imposed by regulatory bodies.


Many organisations have a special categorisation for investments “Regulatory and Mandatory”. These investments are often referred to as

“Must be done!”

A more intelligent description would be:

“Must be done if the organisation wishes to remain in that particular business.”

The reframing is very important. In the late naughties, an investment bank faced with expensive regulatory investments decided to sell its commodities business. This freed up capital that could be invested to make its other business lines stronger.

A common mistake when investing in regulatory and mandatory investments to to try and be the best, to attempt to invest and excel. Using Niel Nickolaisen’s purpose alignment model (Check out Kent McDonald’s excellent description here along with other useful tools), regulatory investments fall in the parity quadrant. Investors should seek to be “good enough” but not invest and excel. This often means implementing a third party solution if an appropriate one is available. Where an appropriate third party solution is not available, the solution should be architected in such a manner that it is easy to migrate to one when it is available.

The strategy for regulatory investments is to minimise total cost of ownership with a “good enough” solution. Unfortunately organisations often misinterpret this as “Implement with your cheapest resources” which is a path to failure and excessive costs. Given that regulations in a market normally increase and are rarely removed, the organisation should consider the long term implications of any solution. This means that regulatory investments should be implemented using eXtreme programming techniques that support safe, rapid and cheap modification in the future.

Realistically, the only way to turn regulatory investments into strategic investments is to deliver a solution to the regulator before your competitors. That way, your organisation can influence the regulators and disrupt any competitors using traditional techniques.

In summary, even though the investment may be obvious, the solution may require careful thought.

Investing with Cynefin: Chaos

To properly understand the Chaos domain in Cynefin, you must consider it from the perspective of the Complex Domain. In the Complex Domain there is enough information for us to form multiple hypotheses. By defintion, in the Chaos Domain, there is not enough information for us to even create a hypothesis. We need to act in order to form hypotheses. Hence:

  • Act
  • Sense
  • Respond

Investments in the Chaos domain should fundamentally be about gaining information.


Novelty often occurs in the Chaos domain because there is a need to act but no known solution. As a result, it is easy to see that innovation occurs in the Chaos domain as no hypotheses exist to guide the investment. Many investments in the Chaos domain are often pure bets where by definition, the pay off is unknowable.

However there is a form of investment in the Chaos domain that is not a pure bet. A form of investment that is often scorned and overlooked… An investment in insight and measurement.

Many organisations invest millions, or even billions in transformation efforts to improve themselves. These organisations will spend a fortune on “consultants” and “thought leaders” but struggle to find funding to demonstrate whether the changes are successful. The reason is that it is impossible to put “accurate” cost saving or revenue estimates on being able to see. Success is often a matter of an executive’s opinion, rather than being based on data. An organisation that cannot measure whether it is successful with an investment will continue to make bad investments, and will fail to learn how to make good ones. The more accurate the accurate the measurement, the more efficient the learning.

Investment in the measurement of the organisation should be overseen by executives to ensure that it happens. An absence of transparency is a failing of the culture, and hence a failing of executives who have a responsibility to investors.

Before investing in a transformation, the organisation should first invest in the measure that will prove the success or otherwise of the transformation. For an Agile transformation, that means investing the measurement of lead time (weighted lead time). This measurement is most easily done on historic data rather than waiting for new data. All that is needed is to link existing data to value rather than functionality. However the value of reorganising existing data is often considered waste and so organisations continue without insight.

Executives who engage on transformations without measurement of lead time time are like drivers who close their eyes whenever they get behind the wheel of a car.

Enabling Constraints and Hippos

One of the most profound insights on enabling constraints came from Marc Burgauer… “Enabling constraints cannot form if a dominant hippo is involved in making the decisions”.


Enabling constraints form agents into a higher order systems. The higher order system provides feedback to the agents to constrain their behaviour and stabilise the enabling constraint.

Part of the process of enabling constraints is the micro-conflicts where the agents give and take in order to align with each other. This will only happen between “peers”. “Peers” means that one of the agents is not a dominant Hippo around whom all the other agents attempt to align. The dominant hippo might be considered the “Apex Predator” in apex predator theory. The apex hippo can dispatch any member of the group without any fear of reprisals. This means that the dominant hippo always wins the prisoner’s dilemma regardless of the outcome. The only way the apex hippo will fail is if the fitness landscape changes, i.e. from a change in the context, or an outside context threat (i.e. Another hippo from the outcome).

Teams and communities that trust each and work in aligned manner do not emerge with an apex hippo present. Instead, identity is formed based on the relationship with the dominant hippo. The goal of each member of the team is to align more closely with the dominant hippo than their colleagues. The goal is to maintain the pecking order. Relationships with other members of the team are incidental and unimportant.

At Skype, we had a apex hippo. Andrew Sinclair was the executive in charge of product. Andrew intuitively understood this and refused to engage in the decision making process. Instead, Andrew used his status and influence to act as a guardian of the process, frequently reminding participants of their responsibilities and the rules of the game, and keeping the group focused on the goal.

The Extreme Tuesday Club in London is the most successful community I’ve been a member of . Simply identifying yourself as a member lead to an immediate trusted relationship with other members. XTC did not have an apex hippo, however XTC had many many leaders to the point where even identifying the leadership was difficult. Conflict was about the only constant at XTC. Conflict and that it always happened on a Tuesday.

I have observed other groups based around an individual or a small group of individuals. Those individuals feel that the community is theirs. None of them evolved into a community where members would bond with each other purely on the basis they were interested in the same thing. More often than not these communities have leaders who are anointed by the apex hippo. They have a revolving door of new members joining as old members leave. They achieve little other than to act as a marketing channel for the apex hippo.

So if you want to build a team, or build a community around an idea, you will need to create an enabling constraint, stimulate micro-conflict, and prevent conflict avoidance.


First kill the apex hippo… especially if it is yourself.