Problem Solving

 Foreword

This book is a guide to help readers learn more effective and efficient ways to solve problems. The format is simple: each concept is explained briefly, followed by small examples to make it easier to understand. In separate sections, there are also complete case examples showing how a problem-solving framework can be applied from beginning to end.

The cases in this book come from a mix of personal fictional examples, fictional cases inspired by consulting work simulations, and my own analysis of real-world problems. The concepts are intentionally explained in a concise way, but still with enough depth to equip readers with practical knowledge they can immediately use in their own problems.

⸻

Chapter 1 — Why Learn Problem Solving?

People often handle problems poorly. Sometimes we jump to conclusions too quickly. Other times we overthink so much that we become unable to act. This is often called analysis paralysis.

Another issue comes from expertise and experience. When someone becomes highly skilled in a field, they may oversimplify problems because they are too used to familiar situations. As a result, they can struggle when facing problems outside their usual experience.

Expertise can also make people unaware of what they do not know. We often think we already see the whole picture, when in reality we only see part of it. This is related to the idea of WYSIATI: What You See Is All There Is.

Because of this, we need structured and general-purpose methods for solving problems.

⸻

Chapter 2 — The Big Picture of the 7 Steps

This problem-solving framework consists of 7 steps:

1. Define the problem

2. Disaggregate

3. Prioritize

4. Workplan

5. Analyze

6. Synthesize

7. Communicate

The cycle is iterative. We can repeat the full cycle anytime as we gain more information or develop a better understanding of the problem and its solution.

The answer we have today is simply a summary of our current understanding. As new information appears, the answer may change.

⸻

Chapter 3 — Define the Problem

The first step is defining the problem clearly.

Some important things to define are:

• What is the main trouble?

• Who owns the problem?

• What does success look like?

• What are the constraints?

• Who are the actors involved?

Trouble

The problem should be written in terms of undesirable outcomes.

Example:

“Store sales dropped by 40% over 3 months.”

This is better than:

“Marketing problem.”

Because it focuses directly on the outcome.

⸻

Owner

Who is most affected by the problem?

The owner matters because the solution should fit their needs.

There may also be important forces around the owner, such as managers, customers, company culture, or economic conditions.

⸻

Success Criteria

Define success in a specific and measurable way.

Bad example:

“Improve the business.”

Better example:

“Increase sales by 20% within 6 months.”

⸻

Constraints

Every problem has limitations.

Examples include:

• time

• budget

• manpower

• legal rules

• technical capability

Example:

“The solution must be implemented within 2 months using less than $500.”

⸻

Refinement

Problem definitions usually need refinement.

Things often adjusted include:

• level of ambition

• boundaries

• level of accuracy

• room for creativity

Sometimes it also helps to challenge the opposite idea.

Instead of asking:

“How do we increase sales?”

Also ask:

“If sales continue to fall, what would most likely cause it?”

⸻

Empathize

For highly human-centered problems, we need to understand people’s experiences and perspectives.

Design thinking is especially useful when building products or services for humans.

⸻

Chapter 4 — Disaggregate

Disaggregation means breaking a large problem into smaller parts so it becomes easier to analyze.

The goal is to find paths toward solutions.

⸻

Logic Trees

Logic trees help us see:

• elements of the problem

• relationships between elements

• levels inside the problem

• important levers

We often create multiple versions of trees to see which one gives the most insight.

⸻

Component Tree

This is the simplest form.

The problem is divided into major factors.

Example for declining sales:

• fewer customers

• lower purchase frequency

• smaller transaction value

Try to make the structure MECE:

• mutually exclusive

• collectively exhaustive

⸻

Deductive and Inductive Trees

Deductive Tree

Used when the logical structure is already known.

Common in mathematics or formula-based problems.

Inductive Tree

Used when we have observations or data but do not yet know the general principle behind them.

In practice, deductive and inductive reasoning are often used together.

⸻

Hypothesis Tree

A hypothesis tree starts with a proposed explanation or solution.

Then we try to confirm or disconfirm it.

Example:

“Sales dropped because customers moved to competitors.”

Then we gather evidence to test whether this is true.

The danger is becoming too narrow and missing better explanations.

⸻

Prioritization

Not every issue has the same impact.

We can prioritize based on:

• potential impact

• our ability to influence it

The Pareto Principle is often useful:

sometimes 20% of causes create 80% of the impact.

⸻

Frameworks

Frameworks help us see problems from certain perspectives.

But every framework contains assumptions. We should not blindly trust frameworks and ignore reality.

⸻

Pitfalls

Some common mistakes:

• confusing correlation with causation

• confusing necessary conditions with sufficient conditions

• becoming too attached to the first hypothesis

⸻

Chapter 5 — Workplan

Analysis should usually begin after forming hypotheses.

Before starting, visualize the output you expect to produce.

⸻

Knock-Out Analysis

Prioritize hypotheses that, if proven wrong, can eliminate many other possibilities.

This saves time and effort.

⸻

One Day Answer

In consulting, teams often create a “one day answer.”

It usually contains:

• the situation

• major observations

• implications or temporary solutions

It is not the final answer, but the current best understanding.

⸻

Strong Hypotheses

Strong hypotheses are easier to test and challenge.

Avoid vague hypotheses.

⸻

Biases

Common biases include:

• confirmation bias

• anchoring bias

• loss aversion

• availability bias

• overoptimism bias

We should remain flexible when facing new data.

⸻

Teams

Teams with healthy discussion norms help reduce biases.

Ideas should be challenged without attacking the people behind them.

⸻

Chapter 6 — Analysis

The purpose of analysis is to understand:

• the scale of the problem

• the direction of influence

• the most important levers

⸻

Heuristics

Some useful thinking tools include:

Occam’s Razor

The simplest explanation is often correct.

Order of Magnitude

Rough estimates are sometimes enough.

Pareto Principle

Focus on the few factors with the largest impact.

Reasoning by Analogy

Learn from similar situations.

⸻

5W1H and RCA

Use questions such as:

• what

• why

• who

• when

• where

• how

Root Cause Analysis helps identify deeper causes.

⸻

Consistency Check

Check whether assumptions are consistent with each other.

⸻

Data Analysis

Use:

• summary statistics

• data visualization

• analysis approach diagrams

For complex problems, visuals are often easier to understand than long text.

⸻

Chapter 7 — Synthesize

Synthesis means combining findings into a coherent story.

In reality, synthesis begins early and improves throughout the process as understanding grows.

Visualizations are very useful in this stage.

⸻

Chapter 8 — Communicate

A good solution is useless if it cannot be explained clearly.

The purpose of communication is to tell the story from problem to solution.

⸻

Leading With the Answer

Sometimes it is best to present the answer immediately.

But for difficult audiences, it may be more effective to explain the reasoning process first.

⸻

Pyramid Principle

The Pyramid Principle helps structure arguments clearly.

Usually:

• the main idea is at the top

• supporting points are underneath

⸻

SCQR

A common structure:

• Situation

• Complication

• Question

• Resolution

⸻

Argument Types

The two main types are:

• deductive

• inductive

⸻

Storyline

Build a clear and compelling storyline.

Good communication helps guide conversations with the problem owner.

⸻

Building the Pyramid

The pyramid can be built:

• top-down

• bottom-up

Depending on the situation and available data.

⸻

Chapter 9 — Uncertainty and Risk

In the real world, we almost never have complete information.

Because of this, problem solving must also consider:

• uncertainty

• risk tolerance

• changing conditions

⸻

Addressing Uncertainty

We must accept that sometimes the best solution is simply a “good enough” solution based on current information.

⸻

Risk Tolerance

Every person and organization has different levels of risk tolerance.

A solution suitable for one party may not fit another.

⸻

Theories of Change

Every solution contains assumptions about how change will happen.

These assumptions should be tested, not blindly trusted.