Critical Thinking & Problem Solving

The company that spent $2 million solving the wrong problem

A mid-size e-commerce company was losing customers. Churn had spiked from 8% to 15% in six months. The VP of Product called an emergency meeting. "Our app is too slow," he declared. "Users are leaving because the experience is frustrating. We need to rebuild the frontend."

The team spent four months and $2 million rebuilding the app. Performance improved dramatically — pages loaded 3x faster. The VP presented the results at the board meeting, confident the churn problem was solved.

Churn went from 15% to 14.5%.

A junior data analyst, who'd been quietly running queries in the background, raised her hand in a team meeting: "I pulled the exit survey data. 73% of churned users cited the same reason: they couldn't find the product they wanted. It's not speed — it's search. Our search algorithm returns irrelevant results."

The team spent three weeks improving search relevance. Churn dropped to 9%.

Four months and $2 million on a speed problem that wasn't the problem. Three weeks and a fraction of the cost on the real issue. The VP's instinct was wrong — not because he was stupid, but because he jumped from symptom (churn) to solution (speed) without diagnosing the actual cause (search).

This is the most common failure mode in business: solving the wrong problem efficiently.

Logical reasoning: the operating system of clear thinking

Before you can solve problems, you need to think clearly. Logical reasoning is the ability to draw valid conclusions from available evidence — and to spot when someone else's reasoning is flawed.

The three types of reasoning:

Cognitive biases: the bugs in your thinking

Your brain is a pattern-matching machine that takes shortcuts to save energy. Most of the time, these shortcuts are helpful. Sometimes they lead you catastrophically wrong. Here are the biases that most damage professional decision-making:

(Percentage of professionals who report being affected by each bias — composite from multiple cognitive bias studies. Treat as directional.)

There Are No Dumb Questions

"If biases are so common, how do smart people still make bad decisions?"

Biases aren't about intelligence — they're about how human brains work. Nobel Prize-winning psychologist Daniel Kahneman, who literally wrote the book on cognitive biases (Thinking, Fast and Slow), said he himself falls for them regularly. The advantage of knowing about biases isn't immunity — it's having a checklist to review before major decisions.

"Can I eliminate my biases?"

No. You can't eliminate them any more than you can eliminate optical illusions. But you can build systems to catch them: checklists, devil's advocates, pre-mortems, and decision frameworks that force you to consider perspectives your instincts miss.

Root cause analysis: the "5 Whys" and beyond

When something goes wrong, most people fix the symptom and move on. Root cause analysis asks: why did the symptom appear in the first place?

The 5 Whys (Toyota Production System):

Ask "Why?" five times to drill past surface symptoms to the underlying cause.

<strong>Problem:</strong> The client received the wrong report.

<strong>Why 1:</strong> Because the analyst sent last month's version. → Why?

<strong>Why 2:</strong> Because the files had the same name and the analyst grabbed the wrong one. → Why?

<strong>Why 3:</strong> Because there's no naming convention or version control for client reports. → Why?

<strong>Why 4:</strong> Because the team has never formalised their file management process. → Why?

<strong>Why 5:</strong> Because the team grew from 2 to 8 people and nobody updated the informal processes. <strong>← Root cause: scaling without process maturity.</strong>

The surface fix is "remind the analyst to check the date." The root cause fix is "implement a file naming convention and version control system." One prevents this incident. The other prevents an entire category of errors.

When 5 Whys isn't enough — the Fishbone Diagram (Ishikawa):

For complex problems with multiple potential causes, organise your analysis by category:

Decision frameworks: structured thinking for high-stakes choices

When stakes are high and information is ambiguous, gut instinct isn't enough. Decision frameworks impose structure on messy situations.

The 2x2 Matrix

The simplest decision tool. Pick two dimensions that matter most, create a grid, and plot your options.

The Pre-Mortem (Gary Klein)

Before making a decision, imagine it's six months from now and the decision was a disaster. Ask: "What went wrong?"

This technique exploits a quirk of human psychology: we're better at explaining past failures than predicting future ones. By projecting into an imagined future where the failure has already happened, people surface risks they'd otherwise suppress.

How to run a pre-mortem:

1. Describe the decision clearly
2. Say: "It's six months from now. We made this decision and it failed spectacularly. What happened?"
3. Everyone writes their answers independently (prevents groupthink)
4. Share and discuss
5. Identify the top 3 risks and build mitigation plans

There Are No Dumb Questions

"When should I trust my gut versus using a framework?"

Trust your gut when: (1) you have deep expertise in the domain, (2) the decision is low-stakes and reversible, and (3) you've made similar decisions many times before. Use a framework when: (1) stakes are high, (2) you're in unfamiliar territory, (3) multiple stakeholders are involved, or (4) you can identify emotional biases influencing your thinking. The best decision-makers use both — gut instinct to generate options, frameworks to evaluate them.

Structured problem solving: the McKinsey approach

Consulting firms have systematised problem solving into a repeatable process. The core framework — used at McKinsey, BCG, and Bain — works for any complex business problem:

Step 1: Define the problem. Write a problem statement that is specific, measurable, and time-bound. "Revenue is down" is not a problem statement. "Q3 revenue in the enterprise segment declined 12% compared to Q2, despite a 15% increase in pipeline" is.

Step 2: Structure the analysis (MECE). Break the problem into mutually exclusive, collectively exhaustive categories. For revenue decline: (a) fewer new deals, (b) smaller deal sizes, (c) lost renewals. These three cover all possibilities without overlapping.

Step 3: Prioritise hypotheses. You can't investigate everything. Which cause is most likely? Start there. "We suspect smaller deal sizes — our average contract value dropped 20%."

Step 4: Analyse and test. Gather data. Talk to customers. Look at the numbers. Does the hypothesis hold up?

Step 5: Synthesise findings. What's the "so what"? Not just "deal sizes dropped" but "deal sizes dropped because we're winning mid-market deals instead of enterprise deals — our positioning has shifted downmarket."

Step 6: Recommend action. Specific, actionable, time-bound. "Reposition our enterprise offering by Q1 with dedicated case studies and a separate pricing page."

Key takeaways

• The most common problem-solving failure is solving the wrong problem. Spend more time on diagnosis before jumping to solutions.
• Know the three types of reasoning — deductive, inductive, and abductive. Most business decisions use abductive reasoning (best available explanation), which requires validation.
• Cognitive biases are universal. You can't eliminate them, but you can build systems to catch them: checklists, pre-mortems, devil's advocates, and data over anecdotes.
• Root cause analysis (5 Whys) prevents you from fixing symptoms. Keep asking "Why?" until you reach the systemic cause.
• The pre-mortem is the most underused decision tool in business. "Imagine it failed — what went wrong?" surfaces risks that optimism suppresses.
• MECE structuring (Mutually Exclusive, Collectively Exhaustive) ensures you don't miss the real cause or double-count.
• Analysis paralysis is real. Most decisions are two-way doors. Decide, act, learn.