The Problem With Problem-Solving

Organisations talk constantly about problem-solving. We want people who can think critically, use data, make good decisions and find solutions quickly. We build systems, processes and technologies to help them do it. Yet we often use the words problem, solution and decision as though they describe the same thing.

They do not.

That distinction matters because good problem-solving starts before the solution and before the decision. It starts with understanding the problem.

A problem is something that requires attention. A solution is how we might address it. A decision is the choice we make about what to do next. These activities are connected, but confusing them can lead us to solve the wrong thing very efficiently.

My research into how practitioners operate in high-performance environments deliberately separated problem type, problem-solving approach and decision-making. The reason was simple: people can face very different problems, use very different ways of solving them and still make fast or slow decisions within that process.

That sounds obvious. In practice, organisations blur these boundaries all the time.

Some problems already have a pathway

Consider a familiar equipment fault. The problem is genuine, but it may be well understood. A diagnostic procedure exists, the likely causes are known and there is an established way of fixing it.

The procedure does not remove the problem. It provides a reliable pathway for solving it.

The same is true of checklists, policies, protocols, algorithms and data thresholds. They capture previous learning so that we do not need to rethink every recurring problem from first principles. Used well, they improve consistency and reduce unnecessary mental effort. My research found exactly this kind of working pattern: experienced practitioners frequently relied on procedures, routines and checklists as part of everyday professional practice.

Within those processes, decisions still need to be made. Does this situation meet the criteria? Which option is appropriate? Is an exception required? Has the process produced the intended result?

This is where decision aids are valuable. They can simplify choices and accelerate action.

But they only work well if we have correctly understood the problem to which they are being applied.

That is the important part.

Familiar does not always mean the same

Experienced people become good at recognising patterns. They see something familiar and quickly connect it to previous knowledge. That is one of the great advantages of expertise.

It is also why phrases such as “we shouldn’t reinvent the wheel”, “we tried that before” and “we’ve always done it this way” are so common.

None of them is inherently wrong.

We should not reinvent things that already work. We should learn from failed attempts. We should preserve effective practices.

The danger comes when precedent replaces enquiry.

“We tried that before” tells us something about a previous solution, but it does not prove that the current problem is the same.

“We’ve always done it this way” explains history, but it does not establish that the assumptions behind the approach still hold.

And we absolutely should avoid reinventing the wheel — once we have established that a wheel is what the problem requires.

That is where effective problem-solving begins to separate itself from simply selecting an available response.

Problems have different natures

Some problems are relatively well-defined. We broadly understand what is happening, credible solutions exist and outcomes can be measured. At the other end of the spectrum are ill-defined problems: causes are uncertain, multiple explanations may be plausible, several parts of the system may interact and no solution is guaranteed.

Both are still problems.

The difference is how much uncertainty exists around what the problem is and how it might be solved.

For a well-defined problem, an established solution, experienced specialist or structured process may be exactly what is needed. For more ambiguous problems, the work changes. We may need to explore competing explanations, understand relationships between causes and consider consequences that are difficult to predict.

This becomes particularly important in organisations because problems rarely sit neatly in one place.

A visible performance issue may actually originate in the system — the processes, workflows or technology through which work is delivered.

It may sit in the structure — accountabilities, reporting lines, resources or the way expertise is organised.

It may be a people issue — behaviour, relationships, trust, communication or capability.

Or it may genuinely be a performance issue — execution, quality or output.

Often, these interact. Systems and structures can themselves create problems, and those problems may emerge both from organisational decisions and from the way work is experienced by the people within them.

This is why the visible symptom is not always the problem.

“Performance is down” describes an outcome.

“The team isn’t collaborating” describes an observation.

“We need better data” suggests a solution.

None of those statements necessarily tells us what is actually wrong.

The problem should shape the response

Once we understand the nature of the problem, we can ask better questions.

Does an established solution already exist? What expertise is relevant? Is that expertise held by one person or spread across several? Do people simply need to coordinate their work, or does the problem require their different perspectives to genuinely interact?

This is where Blended Intelligence begins.

The premise is not that every difficult problem needs a bigger team or more collaboration. In fact, unnecessary collaboration can make problems harder by adding people, relationships and coordination demands.

Some problems need one expert. Some need several people working in parallel. Others require knowledge from different domains to be surfaced, challenged and integrated because no single perspective is sufficient. Adaptive Teaming describes this as matching the configuration of expertise to the demands of the problem rather than treating one form of teamwork as universally superior.

The sequence therefore matters:

The first step is deliberately the hardest to skip.

As technology, data and AI make answers increasingly accessible, generating solutions may become easier. The more important capability may be knowing when an existing answer is appropriate and when the situation requires us to think differently.

Good problem-solving is not about always being innovative, always thinking slowly or constantly challenging established practice. Sometimes the procedure is right. Sometimes experience should allow us to move quickly.

The skill lies in recognising when it does not.

Sometimes we should avoid reinventing the wheel. Sometimes we need to ask whether a wheel is what the problem requires.

It starts with the problem.