My first exposure to structured problem-solving did not come from a business book, a consulting framework or a strategy course.
It came from trying to work out why my computer had stopped working.
My first PC was an old Brother model that I used to play Civilisation and Unreal Tournament: Game of the Year Edition. It would break all the time.
Sometimes it would not start properly. Sometimes a game would stop working. Sometimes the sound disappeared, the internet connection failed or a new piece of software caused something else to go wrong.
There was no AI assistant to ask. Search engines were primitive compared with what we have today, and replacing the computer was not an option.
I had to work it out.
Was it the hardware?
Was it the operating system?
Had something changed since the last time it worked?
Was the issue affecting everything or only one application?
Could I reproduce the problem?
What could I rule out?
Without knowing it, I was building logic trees in my head.
I would break the problem into possible causes, test them one at a time and eliminate branches until I found the most likely explanation.
That process taught me something important from a young age: a complicated problem becomes much easier to solve once it has been divided into smaller, more testable parts.
I first encountered a formal version of this way of thinking in around 2006, when I read The McKinsey Mind.
That was where I discovered MECE: mutually exclusive and collectively exhaustive.
What struck me was not that I had found a completely new way of thinking. It was that somebody had given a name and a structure to something I had already been doing naturally.
Whenever I encountered a complicated problem, my instinct was to break it down.
What are the component parts?
Which issues are connected?
Which issues are genuinely separate?
What would need to be true for the overall outcome to improve?
I had not called it a logic tree. I had not described the branches as mutually exclusive and collectively exhaustive.
I had simply been trying to make difficult problems easier to understand.
Discovering the framework gave me a more disciplined and communicable way to do it.
Nearly two decades later, I still believe structured thinking is one of the most valuable, and most underrated, skills in growth.
Not knowledge of the latest advertising platform.
Not familiarity with another attribution tool.
Not the ability to repeat whichever growth framework is currently fashionable.
The ability to take an unclear commercial problem, structure it properly and work out what is actually happening.
Most growth problems arrive badly defined
Businesses rarely present growth problems in a neat and usable format.
They arrive as statements such as:
- "We need more leads."
- "Marketing is not performing."
- "Our conversion rate is too low."
- "SEO traffic has fallen."
- "Sales need better opportunities."
- "We need to improve retention."
These statements may describe genuine concerns, but they are not yet useful problem definitions.
"We need more leads" could mean the business does not generate enough demand.
It could also mean that it generates plenty of enquiries, but too few match its ideal customer profile.
It could mean sales follow-up is too slow.
It could mean qualification criteria are inconsistent.
It could mean the company has enough opportunities but an unrealistic revenue target.
It could even mean that the business does not need more leads at all. It needs to convert more of the demand it already creates.
The first explanation offered is not necessarily the correct one.
But businesses frequently jump from the original statement to an immediate solution:
- Run more advertising.
- Hire another salesperson.
- Publish more content.
- Redesign the website.
- Implement a new CRM.
This is how companies end up investing heavily in activities that treat the visible symptom without addressing the underlying constraint.
The rush to tactics
Growth teams are surrounded by tactics.
Every channel has specialists, tools, agencies, methodologies and best-practice playbooks attached to it. Each one offers a plausible answer to the growth problem.
The paid media specialist sees an opportunity to increase spend.
The SEO specialist sees a content gap.
The conversion specialist sees friction on the website.
The sales leader sees a lead-quality problem.
The brand consultant sees a positioning problem.
They may all be partially correct.
The difficulty is determining which issue matters most, how the issues relate to one another and where intervention will produce the greatest commercial impact.
That requires more than channel knowledge.
It requires structure.
Without structure, growth strategy becomes a collection of opinions competing for budget.
What MECE actually means
MECE stands for mutually exclusive and collectively exhaustive.
In simple terms, when breaking down a problem, the categories should ideally be separate enough that the same issue does not repeatedly appear across different branches, but complete enough that the branches account for the whole problem.
Barbara Minto helped formalise and popularise this approach through her work on structured thinking and the Pyramid Principle.
MECE is sometimes presented as a rigid consulting technique: a way to make slides appear intelligent or organise information into tidy groups of three.
That misses the point.
The purpose is to improve the quality of the thinking.
It forces you to ask whether your categories genuinely explain the problem, whether they overlap and whether anything important is missing.
It is the difference between listing some possible causes and building a structure that allows you to investigate the problem properly.
Troubleshooting a business is not that different from troubleshooting a computer
The problems are larger and the consequences are greater, but the basic discipline is remarkably similar.
When my old PC stopped working, randomly changing several things at once would have made the problem harder to understand.
If I changed the hardware configuration, reinstalled the software, updated the drivers and adjusted the settings simultaneously, I might have fixed it.
But I would not have known what had fixed it.
Worse, I might have introduced several new problems in the process.
Businesses do this constantly.
Revenue falls and several changes are made at once.
The company launches new advertising campaigns, changes its pricing, redesigns its website, replaces its CRM and restructures the sales team.
Performance may eventually recover, but nobody knows why.
Equally, performance may deteriorate further, and the number of possible explanations has multiplied.
Structured problem-solving encourages a different approach.
- Define the issue.
- Break it into possible causes.
- Identify the evidence needed to test each cause.
- Change one meaningful variable where possible.
- Measure what happens.
- Retain what works and eliminate what does not.
The stakes are different, but the underlying logic remains the same.
Turning a growth problem into a logic tree
Imagine that a company is behind its revenue target.
The immediate reaction may be that marketing needs to generate more leads.
A more structured starting point would be:
Revenue is below target because of one or more of the following:
- The business is acquiring too few customers.
- Customers are spending less than expected.
- Customers are leaving sooner than expected.
That first layer separates acquisition, customer value and retention.
The acquisition branch can then be broken down further:
- The business is generating insufficient demand.
- The demand being generated is low quality.
- Too few qualified prospects become opportunities.
- Too few opportunities become customers.
- The sales cycle is taking longer than forecast.
The demand branch can then be divided again:
- The addressable audience is too small.
- The brand is not reaching enough of that audience.
- The messaging is not creating sufficient interest.
- The offer is not compelling enough.
- The available channels are being used inefficiently.
At this point, "we need more leads" has become a series of testable hypotheses.
That is what makes the approach powerful.
Instead of debating opinions, the team can start looking for evidence.
Good problem-solving begins with diagnosis
A doctor would not normally prescribe treatment based solely on a patient saying, "I do not feel right."
They would ask questions, examine the symptoms, consider possible causes and run tests.
Growth teams are often far less disciplined.
Traffic has fallen, so more content is commissioned.
Conversion has declined, so the website is redesigned.
Pipeline is weak, so paid media budgets are increased.
Revenue is behind target, so marketing is told to produce more leads.
Sometimes those actions work.
But when they do, it can be accidental rather than diagnostic.
Structured thinking creates distance between the problem and the proposed solution.
That distance gives you room to investigate.
Before asking, "What should we do?", it encourages you to ask:
- What exactly is happening?
- Where is it happening?
- When did it begin?
- Which parts of the funnel are affected?
- Which parts are not affected?
- What changed before the problem appeared?
- What evidence would confirm or reject each possible explanation?
The quality of the eventual strategy depends heavily on the quality of those questions.
Logic trees make disagreement more productive
Many growth disagreements are not really disagreements about what the company should do.
They are disagreements about what is causing the problem.
Marketing believes sales are not following up effectively.
Sales believes marketing is generating poor-quality leads.
Product believes both teams are setting the wrong expectations.
Finance believes the cost of acquisition is too high.
Leadership believes the company simply needs to move faster.
Without a shared structure, these positions become departmental arguments.
A logic tree gives the organisation a common model of the problem.
It allows each team to contribute evidence to a specific branch.
- Is lead volume insufficient?
- Is lead quality weak?
- Is contact speed too slow?
- Is opportunity progression poor?
- Is the product failing to meet expectations after the sale?
The conversation becomes less about defending functions and more about testing explanations.
That is particularly important in organisations where departmental tribalism has begun to influence decision-making.
A well-structured problem is much harder to politicise.
Structured thinking exposes missing data
Another benefit of logic trees is that they reveal where the business is operating on assumption rather than evidence.
A team may know how many leads it generates but not how many match its ideal customer profile.
It may know the overall conversion rate but not how conversion differs by channel, audience, service or salesperson.
It may report customer acquisition cost without accounting for gross margin or retention.
It may measure traffic decline without separating lost rankings, lower search demand, technical problems and changes in click behaviour.
The logic tree identifies the questions.
The data should then help answer them.
This is a healthier relationship between strategy and analytics than simply collecting every available metric and hoping that an insight eventually appears.
You do not need more dashboards.
You need to know which questions the dashboards are supposed to answer.
Not every problem fits into a perfect tree
The real world is messy.
Growth problems do not always divide into perfectly separate branches. Brand, demand generation, conversion, pricing, product and retention frequently influence one another.
A pricing change may affect conversion, customer quality and retention.
A positioning problem may reduce advertising performance, sales effectiveness and customer satisfaction.
A product issue may initially appear to be a lead-generation problem because referrals and repeat purchases begin to decline.
MECE should therefore be treated as a discipline rather than a claim of mathematical perfection.
The aim is to make the structure as clear, distinct and complete as reasonably possible.
When branches overlap, acknowledge it.
When information is missing, identify it.
When the structure no longer explains the evidence, change the structure.
The framework exists to support the thinking. The thinking should not be distorted to protect the framework.
The difference between complicated and complex
Structured thinking helps make complicated situations manageable without pretending they are simple.
Simplistic thinking reduces a problem to one convenient explanation:
- "Organic traffic fell because of a Google update."
- "Sales are down because the leads are poor."
- "Paid media is not working because the budget is too low."
- "Customers are leaving because the price is too high."
Structured thinking does something different.
It reduces the problem into smaller components while preserving the relationships between them.
That distinction is important.
The goal is not to remove complexity by ignoring it. The goal is to organise complexity well enough that decisions can be made.
Structure becomes even more important in the AI era
AI has made it dramatically easier to generate possible answers.
It can suggest campaign ideas, analyse datasets, write content, summarise research and propose strategies within seconds.
But generating more answers is not the same as solving the right problem.
If the initial question is poorly defined, AI can help a team travel in the wrong direction much faster.
Ask how to generate more leads and it will provide dozens of tactics.
Ask why qualified pipeline has declined, structure the possible causes and provide the relevant commercial data, and the output becomes considerably more useful.
The value increasingly lies in framing the problem.
AI can assist with research, analysis and execution. Humans still need to determine what should be investigated, which distinctions matter and what evidence is required.
As the cost of producing answers falls, the ability to ask and structure better questions becomes more valuable.
Frameworks should improve instinct, not replace it
I was attracted to MECE because it formalised something that already felt natural to me.
But instinct alone is difficult to communicate.
You might sense that a proposed solution is wrong without being able to explain why.
You might see a connection between several commercial problems that others view separately.
You might recognise that a team is working on symptoms rather than causes.
A framework turns that instinct into something visible.
It enables other people to inspect the logic, challenge the assumptions and contribute additional evidence.
That is what the best frameworks do.
They do not replace judgement.
They make judgement easier to explain, test and improve.
The most underrated growth skill
Growth is often presented as a race to discover the next tactic, channel or technology.
Those things matter.
But the strongest growth leaders I have encountered are rarely defined by their knowledge of one platform.
They are able to enter an unclear situation, identify the real question and give the problem a structure that other people can understand.
They know how to separate evidence from assumption.
They know how to distinguish causes from symptoms.
They know when a problem needs to be broken down further and when additional analysis is no longer useful.
They do not begin with the tactic they want to deploy.
They begin with the problem the business needs to solve.
I began learning that lesson while trying to keep an unreliable old computer running long enough to play another game of Civilisation.
The McKinsey Mind later gave me the terminology.
Years of working in growth demonstrated just how commercially important the underlying skill is.
The most valuable person in the room is not always the person with the fastest answer. Often, it is the person who can structure the question properly.
