Critical Problem Solving: The Human Capability Behind Better Decisions

12 minutes ago

8 min read

For years, organisations have been told that critical thinking will remain important even as technology changes the nature of work. During the rise of Analytics, the argument was that people needed to do more than produce reports; they had to interpret what the data meant. Today, as Generative AI becomes increasingly capable of producing answers, analyses and recommendations, the same concern has become even more important.

But there is a problem with the term critical thinking. It can sound academic and abstract. At its worst, it can even be interpreted as simply questioning or disagreeing with everything.

That is not especially useful to an organisation. Most organisations create value by solving problems — for customers, employees, stakeholders or operations.

So perhaps the more useful capability is what we call Critical Problem Solving.

The goal is not to be critical. The goal is to solve the right problem well.

What do we mean by Critical Problem Solving?

Our working definition is deliberately simple:

Critical Problem Solving is the discipline of solving the right problem, questioning our assumptions, and changing course when the evidence demands it.

In practice, that means being disciplined enough to ask some uncomfortable questions before becoming too attached to a solution:

What are we actually trying to solve, and why does it matter?
What are we assuming about what is causing it?
What biases might be shaping the way we see the problem?
What would success look like?
What would tell us that we are wrong?
And if that evidence appeared, would we actually change course?

This is less about mastering a collection of thinking techniques and more about developing a way of approaching problems and decisions.

Why this matters in the age of Data and AI

We already live in a world awash with data. Organisations have dashboards, reports, surveys and sophisticated Analytics, yet having more data has never guaranteed better insight. It is entirely possible to be drowning in data while still being starved of useful insight.

AI now creates a similar challenge. Generative AI can produce answers, analyses, recommendations and content in seconds. As organisations rush to become “AI-first”, the volume of plausible-looking output will only increase.

But more output does not automatically mean more value.

The real challenge becomes whether we can distinguish:

an important insight from an interesting pattern;
a sound conclusion from a plausible one;
a genuine opportunity for AI from a solution searching for a problem;
evidence that challenges us from evidence we are selectively using to support what we already believe.

In other words:

Can we recognise the gems from the dirt?

Perhaps the scarce capability in an age of abundant Data and AI is not simply producing more analysis or more answers. It is knowing what deserves our attention, confidence and action.

What does Critical Problem Solving look like in practice?

1. Define the problem before jumping to the solution

Many organisational conversations begin with statements such as “We need a dashboard”, “We should use AI” or “We need more training”. But these are usually proposed solutions, not clearly defined problems.

Consider a simple example:

“Sales are falling.” — an observation.
“Our salespeople lack capability.” — an explanation.
“We need sales training.” — a proposed solution.

They are not the same thing.

Critical Problem Solving starts by asking:

What exactly are we trying to solve, and why does it matter?

Until that is clear, it is difficult to know whether the proposed solution makes sense.

2. Separate what is happening from why we think it is happening

Suppose employee turnover rises. Turnover is the observed outcome. Pay, leadership, workload, career opportunities and labour-market conditions may all be possible explanations.

Yet organisational discussions often move quickly from:

“Turnover has increased”

to:

“People are leaving because our salaries are too low.”

The first statement may be supported by evidence. The second is still an interpretation.

Critical Problem Solving asks what else could explain the same outcome. That matters because the explanation we choose determines the solution we are likely to pursue.

Get the explanation wrong, and we may become extremely efficient at solving the wrong problem.

3. Make assumptions and biases visible

Every decision contains assumptions. The difficulty is that we often do not notice them.

Take the statement:

“We should automate this process because AI will make it more efficient.”

That assumes the process is worth automating, that speed matters, that AI can perform the task reliably and that the value created will justify the effort.

Bias can make these assumptions even harder to challenge. We may anchor on the first explanation we hear, give too much weight to recent experience, or selectively notice evidence that supports what we already wanted to do.

Once time, money or reputation has been invested, another problem appears: it becomes psychologically harder to walk away.

Critical Problem Solving therefore asks:

What are we assuming, and what might be influencing the way we see the problem?

The objective is not to eliminate bias completely. It is to recognise where bias might enter the decision and put checks around it.

4. Know what can be tested — and what cannot

Some explanations can be tested relatively easily. If we believe employees are leaving because salaries are uncompetitive, we can look for evidence that supports or weakens that explanation.

But not every organisational problem can be reduced neatly to data. Some are difficult to quantify. Some require proxies. Some involve incomplete information. Others ultimately involve values, priorities or trade-offs where there may be no objectively correct answer.

How much risk should an organisation accept? What is a fair outcome? How much flexibility should employees receive?

Evidence can inform these decisions. It may not settle them.

Critical Problem Solving therefore means recognising:

what can be measured, what can only be approximated, and what ultimately requires judgement.

The goal is not to force every problem into a dataset. It is to understand the limits of what we know.

5. Define success — and failure — before acting

This may be one of the most important disciplines.

Once an organisation commits to a solution, confirmation bias makes it easy to notice evidence that suggests it is working and explain away evidence that does not. Past investment can make this worse; after spending money, time or political capital on an initiative, stopping can feel like admitting failure.

So before acting, Critical Problem Solving asks two questions:

What would success look like?

and equally importantly:

What evidence would tell us that this is not working?

Defining both in advance makes it harder to move the goalposts later. Without that discipline, almost any outcome can eventually be rationalised as success.

6. Change course when reality disagrees with us

Problem solving should not end when the decision is made. Even a well-reasoned solution may not work as expected.

After acting, the organisation should ask:

Did the outcome move in the direction we expected?
Were our assumptions right?
What did we miss?
Are we continuing because the solution is working, or because we are reluctant to abandon it?
What should we change next?

The objective is not to be right the first time. It is to learn quickly enough to make the next decision better.

Problem → Decision → Action → Evidence → Learning → Adjustment

That is how problem solving becomes organisational learning.

One capability, three decision modes

In our previous article, we suggested that organisations broadly make decisions in three ways:

Judgement-led — relying primarily on experience and context.
Evidence-led — using data and analysis to inform the decision.
Automated — handing defined parts of the decision or execution to a system.

Critical Problem Solving sits underneath all three. The underlying discipline remains the same; what changes is where the thinking needs to focus.

In Judgement-led decisions

Sometimes there is little time, little data or no meaningful way to quantify the issue. An experienced leader may simply have to decide.

Critical Problem Solving does not replace judgement. It improves it by asking whether we are solving the right problem, whether past experience really applies to the current situation, what assumptions we are making and what would tell us that our judgement was wrong.

Bias matters here because experience can easily become an anchor. The fact that something worked before does not guarantee that the present situation is the same.

Sometimes, after considering all of this, the answer will still be:

“We cannot know for certain. This is a judgement call.”

That is not a failure of Critical Problem Solving. Recognising uncertainty is part of good judgement.

In Evidence-led decisions

When data is available, Critical Problem Solving focuses less on the mechanics of analysis and more on how we interpret the evidence.

For example, if turnover is higher among younger employees, the pattern may be real. But that does not automatically mean that “young employees are less loyal”. Tenure, job level, pay progression, career opportunities or other factors may explain the same result.

So the thinking questions become:

Are we looking for evidence, or merely looking for confirmation?
What else could explain the same pattern?
Are we treating a convenient proxy as if it were the real outcome?
What would make us change our conclusion?

Analytics may show us what is happening. Critical Problem Solving asks whether we are interpreting it sensibly.

This is also where defining success and failure beforehand matters. If an intervention is introduced, we should know in advance what evidence would convince us that it is not working.

In Automated decisions

With AI and automation, Critical Problem Solving focuses on the reasoning we choose to embed and scale.

Before automating something, we should ask what problem we are solving and why. Are we assuming that faster, cheaper or more consistent automatically means better? What human assumptions are being built into the process? What important context could be lost? And what failure would cause us to redesign or stop?

Consider an organisation using AI to shortlist job applicants. The system may be faster and more consistent, but those benefits do not tell us whether the criteria being automated actually represent what makes someone suitable for the role.

Nor do they tell us whether past hiring decisions contain assumptions or biases worth questioning.

AI can scale a decision. Critical Problem Solving asks whether we should be scaling that decision in the first place.

The same discipline sits across all three

Seen this way, Critical Problem Solving is not another decision mode and it is not another technical skill.

It underpins all three:

In Judgement-led decisions, it challenges our experience.
In Evidence-led decisions, it challenges our interpretation.
In Automated decisions, it challenges the reasoning we embed and scale.

Across all three, the questions remain remarkably similar:

What are we solving? Why does it matter? What are we assuming?What biases might be influencing us? What does success look like? What would tell us that we are wrong? And are we willing to change course?

That is the discipline.

Can Critical Problem Solving really be taught?

A workshop can introduce useful frameworks, questions and techniques. But attending a workshop does not suddenly make someone an expert problem solver.

Someone may understand confirmation bias and still seek evidence that supports a preferred conclusion. Someone may know assumptions should be challenged but hesitate to question a senior colleague. Someone may know AI can make mistakes but accept its output because it appears confident and objective.

So the capability has to be practised. People need repeated opportunities to work through ambiguous situations, expose their reasoning, receive feedback, recognise their biases and learn from outcomes.

It is less like learning a software feature and more like building judgement.

Perhaps this is the human capability that matters more as the tools improve

As Analytics and AI become increasingly accessible, producing information, analyses and recommendations becomes easier. But easier access to answers does not guarantee that we are asking the right questions.

The harder part may increasingly be deciding:

What problem should we solve?
Why does it matter?
What are we assuming?
What biases might be affecting us?
Where does the evidence help — and where does it run out?
What would success look like?
What would prove that we are wrong?
And would we be willing to change course?

That is what we mean by Critical Problem Solving.

It is not criticism for its own sake. It is not endless analysis. And it is not an insistence that every decision must be data-driven.

Critical Problem Solving is the discipline of solving the right problem, questioning our assumptions, and changing course when the evidence demands it.

The goal is not to eliminate judgement, bias or uncertainty. It is to manage them well enough to make better decisions.