Strategy as Enabling Constraints

In one of my recent presentations, I talk about 3 Cs of “Imposing” Agile; Coherence, Constraints and Curiosity. That idea, along with some of the content made it into a whitepaper I wrote last year. A key part of that, and one of the 3 Cs, is the idea of using constraints, and specifically of strategy as a form of constraint. The talk has helped my clarify some of my thinking, and know I want to try and write that down to formalise it some more.

I’ve mentioned my love of running in previous posts, and just before Covid-19 hit us a year ago I was training for a marathon. I like to do one a year, which is enough of a challenge without letting it completely take over my life. Marathon training requires discipline – at least it does for me when I set myself goals and try and beat my Personal Bests (PBs). To maintain the discipline I use a training plan (the picture above) and I like to think of it as a form of constraint which guides my progress through achieving certain paces, times and distances in the lead up to the race. As an example of a constraint, it can be useful to help look at constraints in a number of ways.

To begin with, I could use the training plan in a couple of ways.

1. I could rigidly follow it to the letter, running the prescribed routines, in the prescribed order, on the prescribed days.
2. I could use it as a guide and influence, running similar routines, but with flexibility of order and on days that are convenient to me.

In other words, I could uses the plan in a context free way (ignoring my lifestyle, calendar, health etc.), or a context sensitive way, (allowing for my lifestyle, calendar, health etc.) and each would result in different outcomes from injury and failure to PBs and success. This is how Alicia Juarerro describes constraints. More generally she says that:

constraints constitute changes in probabilities

Lean UX 2015

She uses the example of white noise, which has no constraints and with which every pattern is equally probably. Adding constraints makes different patterns more or less likely. Thus adding, removing or changing constraints will alter the likelihood of different outcomes.

Similarly, Dave Snowden uses the language of governing and enabling constraints.

• Governing constraints are expected to produce a high probability outcome. They are more causal and repeatable, and they limit action regardless of context.
• Enabling constraints are expected to produce a wider range of outcomes, albeit with lower probability. They are more catalytic and probing, and they allow for more freedom of action depending on context.

For my training plan example, I can use the plan to either govern my routine such that I repeat it as prescribed, or to enable my routine such that it helps me choose how I run.

It’s important to note at this point that context free / governing constraints, or context dependant / enabling constraints, are neither good nor bad. Instead it’s more useful to think about what constraints are currently in place, how appropriate they are, and how they could be adjusted. With a training plan some people are strongly motivated by the rigid structure, whereas my preference is to have enough structure to ensure I am making progress, but with the adaptability to fit around my work and home life. If I find I’m not making the progress I would like, however, I might choose to follow the plan more strictly again.

What’s this got to do with strategy? I like the definition of strategy from Henry Mintzberg:

Strategy is a pattern in a stream of decisions

Henry Mintzberg: Patterns in Strategy Formation

This implies that strategy involves people thinking for themselves, which thus means that strategy can be a form of context dependant and enabling constraint. I’ve previously blogged about strategy as choices and using even-over statements.

Further, given Richard Rumelt’s definition of Good Strategy as having Diagnosis, Guiding Policies and Coherent Actions, then the diagnosis is what provides the dependant context, and the guiding policies are the enabling constraints.

If we consider strategies to be context dependant and enabling constraints, then we can think about how to frame strategy by considering how to adjust the constraints in order to allow people the right amount of freedom. For example:

• Are we allowing people to make their own choices or are we making more decisions for them?
• Are we asking more questions or giving more answers?
• Are people having to think for themselves or follow instructions?
• Are we describing an outcome we want to achieve or an activity we want people to perform?
• Are we giving people a direction to move towards or a plan on how to get somewhere?

To reiterate, goal is not to try and contrast good with bad, but describe a tension where we want to find the right balance. If we feel that there is too much freedom, and as a result a lack of clear alignment, then we can consider how we could tighten the constraints so that they are more governing. If we feel that there is not enough freedom and as a result a lack of empowerment or creativity, then we can consider how we loosen the constraints to make them more enabling.

To wrap up and summarise, I would say that thinking about strategy in terms of constraints, and specifically enabling constraints, helps with understanding what a good strategy is and how it can help by providing enough of a boundary to create alignment, but also enough space to allow people to be creative and innovative in the action that they take.