Key concepts for better decisions: path dependence
Key concepts for better decisions: path dependence
Decisions are influenced both by present needs and past decisions. The decisions of the past may push you towards a worse choice. Here's how to identify this and make better decisions.
In a recent session with a CTO, the topic of path dependence came up. It was a concept we often referenced when I worked at SEEK. It's a consideration, particularly when recent acquisitions are involved, as there is often a strong need to consider what currently exists as a result of bringing companies together and what should exist in the future as the organisation consolidates. We regularly referred to this as ‘path dependent vs market dependent’ or variations along these lines.
In its simplest form, it is the idea that historic decisions constrain present decisions because they increase the friction for taking alternative options. This may dampen an organisation’s potential performance because it makes sub-optimal choices. A common example is a product or service implemented on an inferior platform, which reduces the product’s performance or capabilities.
Path dependence has been a key factor in various historical instances, such as the railway gauge differences between Australian states and between the USA and Canada, VHS's triumph over Betamax, QWERTY keyboards' prevailing over DVORAK, and more.
These examples demonstrate how decisions made in the past can significantly impact the present and future, a concept that remains relevant in our decision-making processes today.
It's crucial to remember that market forces are not static; they evolve. Therefore, decisions are always influenced by the pressures of the present situation and the echoes of historical choices, necessitating a comprehensive understanding of both.
Sometimes, it can be helpful to identify when historic decisions may add friction to taking the best path due to natural lock-in. Breaking this lock-in may lead to a shorter route to achieving value sooner or making a larger impact.
Why is path dependence relevant for CTOs?
For CTOs, it's a useful concept because it allows them to assess what choices may have been path-dependent. Simply identifying this helps separate the influence of path dependence from the present need. It also allows them to assess whether options that break out from the established path may lead to higher performance.
While searching for literature on path dependence vs. market dependence, I discovered a notable absence of public resources that effectively distinguished between the two. This is probably because our shorthand of contrasting path dependence with market dependence is oversimplified.
The true opposite of path dependence is path independence. When we think of path independence, we think of something shaped by the present needs of the markets, by the market forces being exerted on the organisation, and absent from the constraints of what has come before.
In addition, while path dependence often refers to large-scale technology decisions or marketplace winners, it’s more often in a social science, strategy, or financial context. A few publications refer to this idea in a product context, and fewer are still targeted at CTOs with examples specific to software product development.
Realising that this idea may not have been deeply explored as a useful public resource, an aim for all my posts, spurred me to delve deeper into the subject. Drawing on my experiences as a CTO, I am exploring how these concepts can be directly applied in our roles, particularly in decision-making processes. I will share how I used this concept in real situations and how outcome-planning approaches can help identify potentially path-dependent options and alternatives.
What are examples of path dependence CTOs may observe?
There are several common examples of path dependence that any CTOs of a decent tenure and scale of organisation are likely to observe:
The choice of platform or toolchain that was selected.
The skillsets and capabilities of the team.
Properties relating to the workforce include type of contract, location, distribution of skills and experience.
Standards adopted.
Degree of commodification and calcification of process and norms.
Here are some examples from my career experience. For a company formed from the merger of a few acquired companies, we ended up with teams spread across a number of locations. Some of these teams had adequate scale such that we could competently support them, and others were sub-scale, where support, such as leadership guidance, coaching, etc., was provided but at a compromised degree compared to the larger locations.
If we were to start the same products and services today, would we select the same number of office locations and the same distribution of people across countries? The answer is no, we wouldn’t, as these were characteristics of our path-dependent choices.
This doesn’t necessarily mean there is now an imperative to change. Whether you change and select a different option depends on the organisation’s current needs and imperatives. Should there be a significant advantage to changing and selecting a path-independent option, by evaluating for path dependence, you are now aware of the strong preference to continue along the path, which may be influencing your judgement.
As I have shared in previous posts, it is very helpful to map out the desired outcomes as best we understand them over different timescales. What I mean by that is that outcomes are closely related to our purpose and might be perpetually relevant. Which are related to our strategy, our currently chosen path for realising our purpose? What outcomes are related to our current focus and so on?
By framing the outcomes we are trying to achieve and making them specific by the evidence we will use to understand our progress, we can establish a frame for thinking about path-dependent options and path-independent options.
The costs may differ significantly, as path-dependent options often have reinforcing aspects that create incentives that add friction to changing the status quo. For instance, the cost to reskill, rehire, or rebuild may seem prohibitive. So moving either to or from open source and so on. I wrote about the preservation of the status quo and some social status examples recently:
With this framing, you can now make an effort to quantify and qualify the factors that may be locking you into a certain path and what it would take to break free from that path. You can make the path dependence explicit and make your peers and stakeholders aware of it so that they can consider it when contributing to decision-making.
Do you have examples of path dependence influencing decisions in your organisation? Have you ever ‘broken out’ from a path-dependent decision? How was this approached? Share your experiences in the comments.
Resources
Plenty has been written about path dependence as it affects business decisions, particularly those relating to technology and technology product choices across history, for instance, this post.
This post by Austin Yang applies a product lens to the concept of path dependence and categorises high-level types of path dependence in a product context.
This paper highlighting path dependency is not yet a theory, as there's not yet irrefutable evidence for why it occurs, and the paper suggests a shift from anecdotal evidence to more formal replicable research:
And ol' faithful, Wikipedia’s definition of path dependence.