Some problems have solutions and others have strategies. Which do you need?


We often don’t know the outcome of our strategies until several months of collecting, monitoring, analyzing and iterating. You can try different forms of A/B tests to be more confident in your decisions, but they aren’t always worth the time or resources. What is worth your utmost attention when leading the design of new strategy and solutions is problem identification. This is your first milestone and where true confidence in future decisions comes from.

Remember that strategies will change – and that’s okay. Welcome it. They will change more often than you might admittedly want them to. There is a certain level of comfort and vulnerability that comes from accepting that there are unknowns. Similar to how you might use Waze to offer you different routes to get from point A to point B during rush hour, problem identification serves a crucial purpose in defining goals and setting the strategies that may lead you to achieve them.

Not all problems require rigorous identification, though.


It’s important to understand that there are different kinds of problems, and that not all require problem identification. A game of chess is a ‘problem’, a printer that has run out of ink is a ‘problem’. In both these cases, the problem is clearly defined and the set of possible actions is understood. This does not necessarily mean that success is easy, understanding the rules of chess does not mean that you are good at it, just that the problem is clearly defined. These are puzzle-problems: cause and affect is understood and success is easily identified. Puzzle-problems can be contrasted with open-ended issues where success exists on a gradient. Take voting. The set of possible choices is sometimes too limited, the connection between your vote and your goals is vague and whether you voted ‘correctly’ is untestable. In this kind of scenario, framing your choices becomes very important for meaningful decision-making.

System problems

Consider now the following: product X is not selling in the US. At first, problem identification may not seem to be an issue. However, upon more careful examination, it becomes apparent that this problem doesn’t provide a causal explanation. There are, it turns out, a number of intertwined explanations as to why X is not selling. It could be that the sales team is ineffective in outreach. It could be that the company’s brand isn’t align with the target market’s values. It could be that we are in the midst of a soul-crushing recession and no one has the money for non-essential products. The crucial point to understand is that the real subject of enquiry is a complex system: the organization that makes and sells product X (1). It may be a helpful metaphor to consider the initial problem statement as a symptom of a disease. Diseases rarely exhibit themselves through a single symptom, nor are symptoms themselves directly the problem in need of address. The goal of a solution designer, similar to a doctor, is to identify a plausible cause and affect explanation for problems, thereby enabling action that targets the ‘sickness’ rather than the ‘symptoms’. Without a causal explanation, a statement such as ‘product X is not selling’ is as useful as a doctor observing ‘the patient is coughing’.

Pure & Applied – Causality – Service Design EdTech

Problem definition is therefore essential. As in medicine, a treatment following a misdiagnosis will be ineffective and may even cause additional harm. Crucially, it is important to be wary of assuming too quickly that a problem is understood. It is entirely possible that sales are low because of the sales team, but to assume this is the case is like a doctor hearing a patient cough in the waiting room and diagnosing them with a sore throat.

The limits of solution-design

We have thus far sought to explain the kind of context where problem identification matters. It’s equally as important to understand some of the limits of solution-design in these kind of contexts. First, many problems are not, strictly speaking, solvable. Sales, for example, can never be ‘solved’, only improved. For companies and organizations with finite resources and time, the act of choosing which problems to solve is itself one of the defining problems that will determine their success. Second, in a complex system such as a company, most any problem will have innumerable causes. The purpose of defining a problem is to identify the most significant causal factors that you can influence (2). Third, be aware that implementing a solution within a complex system can have unintended consequences.

Pure & Applied – Causality Diagram Puzzle versus system – EdTech – Consulting

Problem definition in practice

How can problem definition be improved in practice?

1. Organizations are diverse, so be cautious about implementing stock solutions

One of the key issues that plagues both researchers and problem solvers is external validity. Just because a solution works in one context, it does not follow that it will work in another context. If your organization is seeking to solve an important problem, than it is probably worth taking the time to investigate that problem thoroughly rather than quickly implementing solutions that have worked for other organizations.

2. Build a shared understanding of a problem

This issue is worthy of its own article, but it’s important to remember that organizations are comprised of different groups (e.g. the sales team, the marketing team, etc.). Involve the right teams and stakeholders in problem identification, both because additional perspectives expand the understanding of a problem, and also because collaboration can nurture a company’s culture while failing to consult can cause resentment and conflict when a solution is implemented (i.e. unintended negative consequences).

3. Explore alternative explanations and frames

When trying to identify a problem, it is easy to fall prey to a number of cognitive biases. Some key biases include overconfidence, availability heuristics (placing greater weight on things that come easily to mind), and confirmation bias, among others (furthering the medical analogy, physicians are prone to these same errors: Five pitfalls in decisions about diagnosis and prescribing). One way to break out of these constraints is to purposefully ask both ‘what are explanations that we haven’t explored?’ and ‘what are alternative ways to frame explanations that we have explored?’ An example of the latter is as follows. Tenants in an office building complain about the elevators being slow. The managers of the building ‘solved’ the problem by putting mirrors in the elevators, having reframed the issue from ‘the elevators are slow’ to ‘people are bored waiting in the elevator’ (example drawn from Are you solving the right problems?).

Recommended Readings

This article by HBR echos examples of company problem identification done right.

This article by Deloitte serves as a powerful reminder of the importance of framing questions and removing our own biases for effective problem solving.

This article by Dwayne Spardlin for the HBR was a seminal work in defining problems of the type explored in this article.

* * *


(1) Which exists and sells its products within a larger social-economic system.

(2) In those cases where the most significant causal factors of a problem are beyond your influence, it may be time to move on to a new problem.

This blog was co-authored with D. Ryan Workman, Public policy professional and organizational strategy expert. Frisbee golf enthusiast.