October 22, 2026

1. Learning Objectives

By the end of this module, you will be able to:

1. Break complex problems into smaller, manageable parts.

2. Use ChatGPT to identify possible causes of a problem.

3. Apply structured problem-solving frameworks.

4. Compare multiple options using clear criteria.

5. Identify assumptions, missing information, and potential biases.

6. Evaluate evidence before accepting AI-generated conclusions.

7. Perform basic risk and impact assessments.

8. Use ChatGPT as a decision-support tool without blindly following its recommendations.

2. What Is Problem-Solving?

Problem-solving is the process of identifying a gap between the current situation and a desired outcome, understanding why that gap exists, and selecting an appropriate solution.

For example, imagine a business website receives 1,000 visitors per month but generates very few enquiries.

The problem is not necessarily that the website needs more traffic. Possible causes include:

• Visitors are not the right audience.

• The service offering is unclear.

• The contact form is difficult to use.

• The website loads slowly.

• The call to action is not visible.

• Visitors do not trust the business.

A weak approach is to immediately redesign the entire website.

A stronger approach is to identify the problem, collect relevant evidence, investigate possible causes, and test the most promising solution.

The problem-solving cycle

1. Define the problem.

2. Gather information.

3. Identify possible causes.

4. Develop alternative solutions.

5. Compare the alternatives.

6. Select a solution.

7. Implement the solution.

8. Measure the result.

9. Adjust the approach if necessary.

ChatGPT can assist with every stage, but the quality of its recommendations depends on the information and constraints you provide.

3. Use ChatGPT to Define a Problem Clearly

Many poor decisions happen because the original problem is vague.

Consider these two problem statements:

Vague: "Our marketing is not working."

Specific: "Our website received 1,000 visits last month, but only five visitors submitted an enquiry form. We need to understand the causes of the low enquiry rate."

The second statement is more useful because it describes the current situation and the outcome that needs improvement.

The CLEAR problem framework

Use this practical framework:

• C — Context: What is happening?

• L — Limitation: What constraints exist?

• E — Evidence: What facts or measurements are available?

• A — Aim: What outcome do you want?

• R — Requirements: What conditions must the solution satisfy?

Practical prompt

Act as a problem-solving consultant.

Help me define the following problem clearly.

Context: [DESCRIBE THE SITUATION]

Current results: [PROVIDE AVAILABLE DATA]

Limitations: [BUDGET, TIME, STAFF, OR TECHNICAL LIMITATIONS]

Desired outcome: [EXPLAIN WHAT SUCCESS LOOKS LIKE]

First, rewrite the problem as a specific problem statement.

Then identify:

1. Facts provided

2. Assumptions that have not been verified

3. Missing information

4. Questions that would improve the analysis

Do not recommend a solution until the problem is sufficiently clear.

This prompt helps prevent premature conclusions.

4. Break Complex Problems into Smaller Parts

A complex problem can feel overwhelming when treated as one large issue.

Decomposition means dividing a problem into smaller components that can be investigated separately.

Example: Low sales

Suppose an online business is experiencing declining sales.

Break the problem into five areas:

1. Traffic: Are enough people visiting the website?

2. Audience quality: Are the visitors potential customers?

3. Conversion: Are visitors taking the desired action?

4. Pricing and offering: Does the product meet customer needs?

5. Retention: Are existing customers returning?

Each area requires different evidence.

For example, if traffic is stable but conversion rates have fallen, purchasing more advertising may not address the actual problem.

Prompt: Problem decomposition

Act as a business problem-solving analyst.

Break this problem into smaller components:

Problem: [DESCRIBE THE PROBLEM]

For each component, provide:

• What it means

• Possible causes

• Information needed

• How to investigate it

• Potential solutions

• Risks of acting without enough evidence

Organize the analysis into a clear hierarchy.

Separate known facts from hypotheses. Do not treat possible causes as confirmed explanations.

Use the resulting structure as an investigation plan rather than assuming that every suggested cause is true.

5. Root-Cause Analysis

Root-cause analysis aims to identify the underlying reason a problem occurs rather than treating only its visible symptoms.

For example:

Symptom: Customer complaints have increased.

Possible underlying causes might include unclear product instructions, delivery delays, product defects, or inadequate customer support.

The first explanation that sounds plausible is not necessarily the correct one.

Technique 1: The Five Whys

The Five Whys technique repeatedly asks why a problem occurred.

Illustrative example

Problem: A client project was delivered late.

Why 1: The final work was not ready on the planned date.

Why 2: The review process took longer than expected.

Why 3: Important changes were requested near the deadline.

Why 4: The client had not approved the initial requirements.

Why 5: The project lacked a formal requirements sign-off.

Potential underlying cause: The project workflow did not establish a clear approval checkpoint.

This is only a possible root cause. Real projects may have multiple causes, and the chain must be verified using actual evidence.

Prompt: Five Whys analysis

Act as a root-cause analysis facilitator.

Problem: [DESCRIBE THE OBSERVED PROBLEM]

Conduct a Five Whys analysis.

For each level:

1. State the question.

2. Provide a plausible explanation.

3. Identify evidence that would support or reject it.

Do not assume that every explanation is correct. If the information is insufficient, state what must be investigated.

At the end, distinguish:

• Observed symptom

• Potential contributing factors

• Most plausible root-cause hypotheses

• Recommended verification steps

• Possible corrective actions

Technique 2: Fishbone analysis

A fishbone diagram groups potential causes into categories.

Common categories include:

• People

• Process

• Technology

• Materials or inputs

• Measurement

• Environment

For a delayed website launch, possible causes might include unclear responsibilities, incomplete content, software problems, delayed approvals, or unrealistic scheduling.

Use ChatGPT to generate possible causes, then investigate the ones supported by evidence.

6. Generate Multiple Solutions

A common mistake is selecting the first solution that comes to mind.

A better approach is to develop several alternatives before making a decision.

Suppose a training institute wants to increase student enrolment.

Potential approaches include:

1. Improve the website and course information.

2. Publish more educational content.

3. Run a targeted advertising campaign.

4. Organize a free demonstration class.

5. Improve follow-up with existing enquiries.

6. Develop partnerships with local organizations.

Each approach has different costs, benefits, and implementation requirements.

Prompt: Solution generation

Act as a strategic problem-solving consultant.

Problem: [DESCRIBE THE PROBLEM]

Available resources: [BUDGET, TEAM, TIME, TOOLS]

Constraints: [LIST IMPORTANT LIMITATIONS]

Generate at least five possible solutions.

For each solution, explain:

• How it addresses the problem

• Expected benefits

• Required resources

• Implementation difficulty

• Possible risks

• Evidence needed to evaluate success

Include at least one low-cost option and one option that can be tested on a small scale.

Do not claim that a solution will work without supporting evidence.

The purpose is to expand the range of choices before evaluating them.

7. Compare Options with a Decision Matrix

A decision matrix helps compare alternatives using consistent criteria.

Imagine a business wants to choose a way to generate new customer enquiries.

The available options are:

• Option A: Improve organic content

• Option B: Run paid advertising

• Option C: Organize a free webinar

The team may evaluate each option on cost, speed, expected reach, and ease of implementation.

Example scoring system

Use a score from 1 to 5, where 1 is poor and 5 is excellent.

For this example, assume that a higher score is better for every criterion.

The weighted score is calculated as:

Weighted score = Sum of (criterion score × criterion weight)

Example for organic content:

[
(5 \times 0.25)+(2 \times 0.25)+(3 \times 0.25)+(4 \times 0.25)=3.5
]

The illustrative scores are:

• Organic content: 3.5 out of 5

• Paid ads: 4.0 out of 5

• Free webinar: 3.0 out of 5

Under these assumptions, paid advertising has the highest score. However, the scores are subjective estimates, not actual performance data.

A different budget, target audience, or business objective could produce a different result.

Prompt: Decision matrix

Act as a decision-analysis consultant.

Decision to make: [DESCRIBE THE DECISION]

Options: [LIST OPTIONS]

Constraints: [BUDGET, DEADLINE, CAPACITY]

Create a weighted decision matrix.

Requirements:

1. Recommend relevant evaluation criteria.

2. Suggest weights that add up to 100%.

3. Explain the scoring scale.

4. Score each option only when sufficient information is available.

5. Label estimates and assumptions clearly.

6. Calculate weighted totals.

7. Explain the strongest advantages and disadvantages of each option.

8. Perform a sensitivity check by identifying which changes in weights could alter the ranking.

Do not present subjective scores as verified facts. Identify what real-world information should be collected before committing to a decision.

8. Critical Thinking: Separate Facts from Assumptions

Critical thinking means examining claims, evidence, logic, and alternative explanations before accepting a conclusion.

AI-generated text can sound confident even when the underlying information is incomplete.

For example:

Claim: "Our social media posts are receiving fewer enquiries because the posting time is wrong."

This may be true, but other explanations are possible:

• The audience has changed.

• The content is less relevant.

• The offer is unclear.

• The call to action is weak.

• Website tracking is not working.

• Overall demand has changed.

The claim should be treated as a hypothesis until evidence supports it.

The fact–assumption–hypothesis distinction

Fact: A statement directly supported by reliable information.

Assumption: Something treated as true for the purpose of analysis but not yet verified.

Hypothesis: A proposed explanation that can be tested.

Inference: A conclusion drawn from available information.

Keeping these categories separate improves decision quality.

Prompt: Critical review

Act as a critical-thinking analyst.

Evaluate the following claim:

Claim: [INSERT CLAIM]

Available evidence: [PROVIDE EVIDENCE]

Analyze:

1. What is directly supported by the evidence?

2. What assumptions are being made?

3. What information is missing?

4. What alternative explanations exist?

5. What evidence would disprove the claim?

6. What is the strongest reasonable conclusion?

Avoid agreeing automatically. Be fair to both supporting and opposing evidence.

Clearly state the limits of the analysis.

This is especially useful before making business, academic, or project decisions.

9. Challenge Your Own Ideas

People naturally tend to favor information that supports their existing beliefs. This tendency is often called confirmation bias.

ChatGPT can help challenge an idea by identifying counterarguments and alternative explanations.

Suppose you believe that launching a new service will automatically increase company revenue.

Ask ChatGPT to examine:

• Whether the target market exists

• Whether customers are willing to pay

• Whether the business has the capacity to deliver

• Whether the service competes with existing offerings

• What evidence would justify a small pilot

Prompt: Devil's advocate

Act as a constructive devil's advocate.

My proposal: [DESCRIBE THE IDEA]

My reasons for supporting it: [LIST REASONS]

Identify the strongest reasonable objections to this proposal.

For each objection:

• Explain the concern

• Identify evidence that would support it

• Suggest how the concern could be tested

• Explain whether the risk can be reduced

Also identify the proposal's strongest arguments in favor.

Do not criticize the idea merely for the sake of criticism. Aim for a balanced evaluation.

A useful decision process considers both the strongest case for an idea and the strongest credible case against it.

10. Evaluate Evidence and Source Quality

When a decision depends on external facts, use reliable evidence rather than relying only on a generated answer.

A practical source-evaluation checklist includes:

1. Authority: Who created the information?

2. Accuracy: Can the claim be independently verified?

3. Relevance: Does the information apply to the decision?

4. Recency: Is the information current enough?

5. Methodology: How was the information collected?

6. Limitations: What does the source not establish?

For example, a marketing report may show that enquiries increased after a campaign. That does not automatically prove that the campaign caused the increase. Other changes may have contributed.

When appropriate, compare the timing, audience, baseline performance, and other plausible explanations.

Prompt: Evidence evaluation

Act as an evidence-evaluation analyst.

Decision or claim: [DESCRIBE IT]

Sources or evidence: [PASTE INFORMATION OR PROVIDE SOURCE DETAILS]

Evaluate the evidence using authority, accuracy, relevance, recency, methodology, and limitations.

Separate:

• Verified information

• Claims that need verification

• Inferences

• Unanswered questions

If the supplied evidence is insufficient, state that clearly. Do not invent sources, quotations, or statistics.

For important current facts, consult primary sources or other credible, independently verifiable references.

11. Risk Assessment and Contingency Planning

A solution may offer benefits while introducing new risks.

Risk assessment helps you identify what could go wrong, how serious the consequences might be, and what can be done to reduce the impact.

Example: Launching a new website

Possible risks include:

• Technical failures

• Incorrect pricing or service information

• Broken enquiry forms

• Poor mobile performance

• Security weaknesses

• Delays in receiving approvals

A simple risk register can organize these concerns.

These ratings are illustrative. A real assessment should reflect the project's actual conditions.

Prompt: Risk assessment

Act as a project risk analyst.

Project or decision: [DESCRIBE IT]

Available resources and constraints: [ADD DETAILS]

Identify potential risks and create a risk register with:

• Risk description

• Likely causes

• Likelihood

• Impact

• Early warning signs

• Preventive action

• Contingency plan

• Responsible person, if known

Clearly label uncertain ratings as preliminary estimates.

Include risks caused by missing information, technical failures, human error, and external dependencies.

Do not invent project details that have not been supplied.

12. Decision-Making Under Uncertainty

Not every decision can be made with complete information.

The important question is whether you have enough evidence to make a reasonable decision, or whether you should collect more information first.

Consider a business deciding whether to invest in a new service.

You may have three options:

1. Launch the full service immediately.

2. Run a small pilot with a limited budget.

3. Delay the decision while gathering more evidence.

The best option depends on the cost of being wrong, the cost of waiting, the available evidence, and whether the decision can be reversed.

Reversible and difficult-to-reverse decisions

More reversible decisions: Testing a new social media format or trying a small change to a landing page.

More difficult-to-reverse decisions: Signing a long-term contract, making a major capital investment, or committing to a large operational expansion.

A reversible decision can often be tested quickly. A difficult-to-reverse decision usually deserves stronger evidence and more careful review.

Prompt: Decision under uncertainty

Act as a decision-support analyst.

Decision: [DESCRIBE THE DECISION]

Available options: [LIST OPTIONS]

Known facts: [LIST VERIFIED INFORMATION]

Unknowns: [LIST MISSING INFORMATION]

Consequences of a wrong decision: [DESCRIBE RISKS]

Compare the options using:

1. Expected benefits

2. Potential losses

3. Reversibility

4. Cost of waiting

5. Information that could change the decision

6. Possibility of running a small test first

Recommend a practical next step, but distinguish evidence-based conclusions from assumptions.

Identify any decision that requires qualified professional advice or authorized human approval.

13. Use Experiments to Test Ideas

Instead of debating endlessly about which solution might work, test the most important assumptions when it is practical and safe to do so.

An experiment should define:

• The hypothesis

• The change being tested

• The target audience or sample

• The measurement

• The duration

• The success criteria

• The limitations

Example: Testing a call to action

Hypothesis: A clearer call-to-action button will increase enquiry submissions.

Test: Compare two versions of a landing page while keeping other important factors as consistent as possible.

Measurement: Enquiry submissions divided by eligible visits.

Important limitation: A small sample or simultaneous marketing changes may make the result inconclusive.

Prompt: Experiment design

Act as an experiment-design consultant.

Business question: [DESCRIBE THE QUESTION]

Hypothesis: [STATE THE EXPECTED EFFECT]

Available resources: [BUDGET, TIME, DATA]

Design a practical test with:

1. Testable hypothesis

2. Independent variable

3. Outcome metric

4. Comparison or baseline

5. Data collection plan

6. Success criteria

7. Potential confounding factors

8. Ethical or operational constraints

9. Decision rule after the test

Do not invent sample sizes or claim statistical significance without an appropriate basis. Explain when the test may be too small to support a confident conclusion.

Testing helps convert opinions into evidence-informed decisions.

14. Use ChatGPT to Create an Action Plan

A good decision is not complete until it can be translated into action.

An action plan should define the tasks, priorities, dependencies, owners, deadlines, and measures of success.

Prompt: Decision to execution

Act as a project execution consultant.

Chosen solution: [DESCRIBE THE SOLUTION]

Objective: [DEFINE SUCCESS]

Available resources: [LIST RESOURCES]

Deadline: [ADD DEADLINE]

Create an action plan containing:

• Task

• Priority

• Dependencies

• Estimated effort

• Owner, if specified

• Deadline, if specified

• Success metric

• Risk or blocker

Do not invent assigned owners or confirmed deadlines.

Identify the first three practical actions and explain how progress should be reviewed.

Once the plan is created, review progress against the actual results and update the plan as circumstances change.

15. Common Problem-Solving Mistakes

Mistake 1: Accepting the first explanation

Problem: The first explanation may be plausible but incorrect.

Improvement: Generate alternative hypotheses and identify evidence that distinguishes them.

Mistake 2: Confusing correlation with causation

Problem: Two events occurring together does not prove that one caused the other.

Improvement: Consider alternative causes, timing, and suitable test designs.

Mistake 3: Using subjective scores as facts

Problem: A decision matrix can appear precise even when its scores are guesses.

Improvement: Label estimates, justify weights, and test whether the ranking changes under different assumptions.

Mistake 4: Providing insufficient context

Problem: ChatGPT may fill gaps with generic assumptions.

Improvement: Provide verified facts, constraints, objectives, and missing-information notes.

Mistake 5: Ignoring implementation constraints

Problem: A technically attractive solution may be too expensive or difficult to execute.

Improvement: Evaluate resources, dependencies, risks, and the cost of implementation.

Mistake 6: Treating ChatGPT as the final authority

Problem: AI-generated conclusions may contain errors or overlook important context.

Improvement: Independently verify important claims and keep consequential decisions under appropriate human control.

16. Practical Project: Solve a Real Business Problem

Project objective

Use ChatGPT to investigate a business problem and create an evidence-informed action plan.

Scenario

A small training institute has received fewer course enquiries over the last two months. The institute is considering spending more money on advertising, but it has not yet identified the cause of the decline.

Your task

Use ChatGPT to investigate the problem systematically.

Step 1: Define the problem

Write a clear problem statement and list the information you need.

Step 2: Identify hypotheses

Possible explanations might include changes in advertising reach, audience quality, course demand, website performance, or follow-up practices.

Treat these as hypotheses, not established facts.

Step 3: Create an investigation plan

Identify the data needed to evaluate each hypothesis.

Step 4: Generate alternatives

Consider several solutions, including low-cost options and small-scale tests.

Step 5: Compare options

Use a decision matrix with clearly explained criteria and weights.

Step 6: Assess risk

Identify the risks of increasing advertising spend before diagnosing the decline.

Step 7: Develop an action plan

Create prioritized tasks, metrics, and review points.

Complete project prompt

Act as a business problem-solving consultant for a small training institute.

Problem: Course enquiries have declined over the last two months. The cause is unknown.

Constraint: The institute has a limited marketing budget and wants to avoid unnecessary spending.

Develop an evidence-informed investigation and action plan.

Include:

1. A precise problem statement

2. Facts known and information missing

3. At least five possible hypotheses

4. Data needed to test each hypothesis

5. A practical investigation sequence

6. At least four possible interventions

7. A weighted decision matrix template

8. A risk register

9. A small-scale testing plan

10. A 30-day action plan

11. Key performance indicators

12. Review and adjustment rules

Do not assume the reason for the decline is known. Do not invent historical figures or claim that a specific marketing intervention will succeed. Clearly distinguish hypotheses, estimates, and verified findings.

Project success criteria

Your completed project should demonstrate that you can:

• Define a problem precisely

• Separate facts from assumptions

• Investigate root causes

• Compare solutions consistently

• Consider risks and uncertainty

• Recommend measurable next steps

17. Knowledge Check

Multiple-choice questions

Q1. What should you do before recommending a solution?

A. Select the most popular solution
B. Define the problem and understand the available evidence
C. Increase the budget
D. Ignore constraints

Q2. What is the main purpose of root-cause analysis?

A. Identify possible underlying causes of a problem
B. Assign blame immediately
C. Create more reports
D. Avoid collecting evidence

Q3. What does a decision matrix help you do?

A. Guarantee a successful decision
B. Compare alternatives using explicit criteria
C. Eliminate uncertainty completely
D. Replace all professional judgment

Q4. What is confirmation bias?

A. Checking every claim against multiple sources
B. Favoring information that supports existing beliefs
C. Recording a meeting
D. Comparing implementation costs

Q5. Which statement best describes a hypothesis?

A. A proposed explanation that can be tested
B. A fact that never changes
C. A guaranteed result
D. A confirmed deadline

Q6. Why should decision-matrix scores be labeled carefully?

A. They are always wrong
B. They can include subjective estimates and assumptions
C. They cannot be calculated
D. They remove the need for data

Q7. What should you consider when making a decision under uncertainty?

A. Only the potential benefits
B. Only the easiest option
C. Benefits, risks, missing information, and reversibility
D. Only the first recommendation

Q8. What is the purpose of testing a solution on a small scale?

A. To avoid measuring results
B. To gather evidence before making a larger commitment
C. To guarantee success
D. To eliminate all implementation costs

Practical challenge

Choose a real but manageable problem from your work, study, or a personal project.

Prepare a one-page analysis containing:

1. Problem statement

2. Verified facts

3. Assumptions and hypotheses

4. Missing information

5. Alternative solutions

6. Decision criteria

7. Risks

8. Recommended next step

9. Success metric

Do not include sensitive personal or confidential business information in the exercise.

18. Answer Key

• Q1: B — Define the problem and understand the evidence.

• Q2: A — Investigate possible underlying causes.

• Q3: B — Compare alternatives using explicit criteria.

• Q4: B — Favoring information that supports existing beliefs.

• Q5: A — A proposed explanation that can be tested.

• Q6: B — Scores may depend on assumptions and subjective judgments.

• Q7: C — Evaluate benefits, risks, missing information, and reversibility.

• Q8: B — Gather evidence before making a larger commitment.

Practical challenge evaluation

A strong analysis clearly separates verified facts from assumptions, considers multiple explanations, defines how evidence will be collected, compares realistic alternatives, and identifies measurable next steps.

19. Module Summary

In this module, you learned how to use ChatGPT as a structured problem-solving and decision-support assistant.

Key takeaways:

• Define the problem before selecting a solution.

• Break complex problems into manageable components.

• Use Five Whys and fishbone analysis to investigate possible causes.

• Generate alternatives before making a decision.

• Compare options using explicit criteria and weighted decision matrices.

• Separate facts, assumptions, hypotheses, and inferences.

• Challenge your own ideas and consider alternative explanations.

• Evaluate source quality and the limits of available evidence.

• Assess risks and uncertainty before committing resources.

• Test promising solutions and measure the results.

• Convert decisions into practical action plans.

• Verify important AI-generated conclusions before acting on them.

Core principle: ChatGPT can help structure your thinking, but reliable decisions require evidence, sound judgment, and accountability.