Tools for Risk: Scenarios
Humans tend to be not very good at risk assessment. There are a lot of reasons for this. Popular discussion of risk tends to be imprecise, with the word "risk" being used in different contexts, some positive, some negative, and often confused with chance, peril, or investment. Human physical tools to assess risk are quite varied, with decision-making in the prefontal cortex often balanced with the amygdala and the limbic system – sometimes subconsciously and very quickly, for better or worse. And the conceptual tools to manage risk are often complex and hard for laymen to apply consistently.
Consider the major decisions in life: where to obtain an education, who to marry, what career to pursue, whether to have children, how to manage finances. Now consider seemingly less important decisions: which car to buy, what color paint to use for the kitchen, whether to plant hibiscus or begonia by the garage. Is our approach and time spent consistent with their impacts on outcome? If you are like most people, probably not. Many outsource the big decisions, particularly in finance, and obsess over minutiae on the small things. My wife accused me of spending hours asking about features of our last television just to bore her to the point where she ceded the decision to me. She had a point.
One method that seems to work well, borrowed from management consulting, is scenario planning and the simple 2x2 matrix. There are a few variants of this Christian and I have come across on this.
One is the sorting matrix, where you consider two dimensions that matter to you and have negative or positive outcomes. IT advisory firm Gartner has popularized this in its Magic Quadrant where potential suppliers of a product or service are rated according to completeness of vision ("can the tool do the work") and ability to execute ("can you implement the tool easily"). There are many variations of this kind of rating system, but they all are biased towards the upper right – preferred alternatives rate highly on both axes. It's a simple way to sort options into manageable groups for consideration.
Selection of the right axes, of course, is important. It is easy to confuse the map for the terrain in this space. "Cost" and "Benefit" are simple and easy to understand, but they tend to mask underlying issues with benefits attainment. A lot of strategy – which is after all, managing risk to produce a desired outcome – comes down to making sure you ask the right questions.
Disruption Theory is a good examples of the pitfalls in analyzing the wrong risks and thereby making poor decisions based on correct, but irrelevant, data. One classic example of this is of course Blackberry, which held 50% market share in the smartphone market and now is defunct in that market. During the run up to the disastrous Playbook launch, Blackberry was growing revenue and all appeared well. But sharp observers noted the average price per device was falling – Blackberry was driving revenue at the expense of burning through potential customers. When the iPad and Playbook arrived, Blackberry's real need – technical parity – became obvious. Their stock dropped 80% over the following year and they never recovered.
An alternate way to sort potential risks and outcomes is scenario planning, where there is not a bias towards negative and positive outcomes. Rather, outcomes are evaluated among two unrelated dimensions, and your plan is tested against the four different outcomes implied. Christian and I wrote about this in our Insurance 2025 paper, which has held up pretty well since we wrote it in 2017. We chose two technology dimensions that would affect the insurance market: AI prevalence and computing centralization – and then considered what an insurance would look like across permutations of those trends.
One end result of scenario planning is "no-regret moves". Looking across the scenarios you are considering, are there actions that can be taken that will be positive across the scenarios? These are actions that can be taken risk-free. There is much to be said for making progress regardless of conditions.
No-regret moves bring their own risk though – they won't defend against the disruption that Blackberry faced. Looking at scenarios will help reduce risk, but they will not eliminate it. We'll explore how to reduce bias in decision-making beyond "cover all the bases" in future posts.