5 Seven Essential Prompt Techniques for Business Teaching

The right technique is not the cleverest one. It is the one that matches the kind of thinking you want your students to do.

5.1 Why These Seven?

These seven prompting patterns offer different ways to organise practice. Their educational value depends on the task, evidence, feedback and learner activity; the list is not a claim that each pattern has been validated across every discipline.

Reverse Prompting: teaches comprehensive scoping and requirement gathering
Pros and Cons: builds analytical decision-making skills
Stepwise Tutoring: reinforces process adherence and documentation
Role Play: develops communication and interpersonal skills
Debating: strengthens strategic thinking and ethical reasoning
Formative Assessment Generator: supplies draft practice questions and provisional feedback
The Expert Panel: develops multi-perspective analysis and synthesis skills

Each pattern needs adaptation to the discipline, supplied records and intended capability. The examples below use different disciplines to show the range, so adapt the prompts for your own field.

These seven are adapted from Conversation, Not Delegation, which covers the same approaches for a general audience. If you want the underlying rationale for why each technique works, see the companion book.

5.2 Technique 1: Reverse Prompting (for Scoping)

Instead of solving a problem immediately, the AI asks the student questions to help scope out all requirements, considerations, and potential issues. This mirrors professional work where defining the problem is half the battle.

Example (HR: policy design):

I need to draft a new "Flexible Work
Arrangement" policy for a 500-employee company
with both office-based and remote staff.

Your task: Ask me a series of questions to help
me clarify all the requirements, considerations,
and potential pitfalls for this policy.

Ask one question at a time. Wait for my response
before asking the next. Continue until you've
helped me think through at least 10 different
aspects.

Begin with your first question.

The AI asks questions the student must think through before responding: fairness, logistics, legal compliance, technology needs, cultural impact. The questions may broaden their scope, but cannot guarantee completeness. Compare the resulting scope with the supplied task criteria.

Adapt for your discipline: Replace the policy topic with whatever scoping challenge fits: a market entry strategy, an audit plan, a supply chain redesign, an IT implementation. The technique is the same: AI questions, student thinks.

Teaching tip: Use as a pre-writing exercise. Ask for a short evidence index and the subsequent draft. Assess the reasons for consequential decisions; a complete private chat history is not required.

5.3 Technique 2: Pros and Cons (for Decision Making)

The AI systematically analyses multiple approaches to a problem, evaluating each against specific criteria. Students must then critically evaluate the analysis rather than just accept it.

Example (Supply Chain: strategy evaluation):

Compare different supply chain strategies for a
mid-sized manufacturer: make-to-stock,
make-to-order, and mass customisation.

For each strategy, provide:
1. A brief description of how it works
2. Three key advantages
3. Three key disadvantages

Evaluate each specifically in terms of:
- Cost efficiency
- Responsiveness to demand changes
- Inventory risk

Conclude with a recommendation for a company
facing unpredictable seasonal demand and justify
your choice.

The follow-up task: Check one consequential advantage and one disadvantage against supplied evidence. Identify a genuine omission if there is one. Students may retain the recommendation when it is justified; do not require a different strategy merely to demonstrate opposition.

Teaching tip: Project the AI’s analysis on screen. Divide students into groups, each arguing for a different strategy. They use the AI’s framework but add their own reasoning.

5.4 Technique 3: Stepwise Tutoring (for Process)

The AI walks through a complex process one step at a time, pausing after each step until the student signals readiness. Use a checked teaching procedure and inspect each learner attempt.

Example (Accounting: audit process):

I am learning how to conduct an audit engagement
from planning through to the audit opinion.

Walk me through the entire process. For each
step, tell me:
1. What action to take
2. What to document
3. What professional standard or legal
   requirement applies at this stage

Use the supplied teaching procedure and criteria.
Do not invent standards or legal requirements.
After each step, ask me to apply it to the
supplied case and justify my decision. STOP
while I answer. Identify missing information.
Do not provide the entire process at once.

Begin with Step 1.

Typing “next” would not demonstrate understanding. Inspect the student’s application and justification against checked teaching criteria before proceeding. A generated explanation is not a view into the model’s internal reasoning.

Teaching tip: Have students repeat the exercise with complications introduced at each step: “What if the client refuses to provide documentation at Step 3?” Check any generated guidance against the supplied procedure before relying on it.

5.5 Technique 4: Role Play (for Communication Skills)

The AI adopts a specific persona and engages in a realistic conversation. This is the “flight simulator” concept: students practice difficult professional interactions where the AI responds dynamically.

Example (Management: difficult conversation):

Play Alex in the fictional P1 practice pack.
Use only P1-D1 to P1-D5.
Open with the concern about dependencies in P1-D4.
I am the junior adviser, not the decision-maker.
Do not invent private history, policy or evidence.
Say when an answer is not supplied.
Stop immediately on STOP or after six exchanges.
Do not require disclosure, agreement or a final ruling.
I will respond to your opening statement.

The learner practises wording, evidence checks and a bounded next step. Text role-play does not establish competence in every aspect of a live workplace meeting.

Adapt for your discipline: AI plays a demanding customer (marketing), a resistant employee (management), a skeptical auditor (accounting), a supplier announcing delays (supply chain), a stakeholder with unclear requirements (IT).

Teaching tip: Ask for a relevant excerpt or the response to the fixed opening, with a short explanation of what the learner would retain or change.

5.6 Technique 5: Debating (for Strategic and Ethical Analysis)

The AI examines multiple perspectives on a contentious issue, argues both sides, and helps students see complexity and trade-offs.

Example (Economics: multi-stakeholder debate):

Simulate a leadership debate on whether to
implement a carbon tax.

Create three personas who will each argue from
their perspective:

Persona 1: Environmental Economist
Focus: climate impact, externalities, long-term
economic benefit

Persona 2: Industry Representative
Focus: competitiveness, cost burden, employment
impact

Persona 3: Treasury Official
Focus: revenue, fiscal policy, implementation
feasibility

Have each persona make their opening argument.
Use only the supplied teaching sources. Label these
as generated perspectives, not independent evidence.
Do not invent empirical claims or citations.
Then have them respond to each other through
three rounds. After the debate, summarise the
key tensions and what a compromise policy might
look like.

Begin the debate.

Students cannot just pick a side. They must understand legitimate competing perspectives and make a justified recommendation that acknowledges trade-offs.

Teaching tip: This can support practice in analysing a case. Instead of memorising model answers, they practise analysing competing priorities.

5.7 Technique 6: Formative Assessment Generator (for Practice)

An educator can request draft questions, scenarios or quizzes for low-stakes practice. Review the question, intended answer and feedback before student use.

Example (Tourism & Hospitality: scenario practice):

I'm studying for an exam on service recovery in
hospitality management.

Generate 10 short scenarios (2-3 sentences each)
where I need to identify:
1. What went wrong in the service delivery
2. The appropriate recovery strategy
3. A possible impact on guest loyalty and what
   evidence would be needed to investigate it

Use the supplied teaching criteria and answer notes.
After each attempt, offer provisional feedback
with its source. Flag ambiguity instead of
inventing a definitive answer.

Make the scenarios progressively more complex.

Begin with Scenario 1.

Students can request further practice, subject to access and tool limits. Use checked questions and feedback, then an independent changed problem to examine learning. Repetition and a model’s approval do not establish mastery.

Critical principle: This is for practice, not graded assessment. Students need to struggle and make mistakes in low-stakes environments.

5.8 Technique 7: The Expert Panel (for Multi-Perspective Analysis)

Students consult multiple AI “experts” with different professional perspectives on the same problem, then synthesise the competing advice into a strategic recommendation.

Example (Information Systems: system implementation):

Our company needs to decide on an ERP
implementation approach. Create three expert
personas:

Expert 1: IT Architect
Focus: technical feasibility, integration,
security, scalability

Expert 2: Change Management Consultant
Focus: user adoption, training, organisational
readiness, resistance

Expert 3: CFO
Focus: cost, ROI, risk, business case

Treat these as generated viewpoints, not real
experts or independent evidence. For each provide:
1. Their assessment of the key risks
2. Their recommended approach
3. Limitations of their recommendation
4. What data they would need to validate it

Begin with Expert 1.

After receiving the perspectives, students compare the claims, check them against supplied records or independent sources, and justify a recommendation. Agreement among personas is not corroboration.

Teaching tip: This technique combines well with Debating (have the experts debate each other) and Role Play (present your synthesis to a skeptical stakeholder).

5.9 How to Choose Which Technique to Use

If you want students to…	Use this technique
Define a complex problem comprehensively	Reverse Prompting
Evaluate competing options and justify a choice	Pros and Cons
Follow a sensitive process correctly	Stepwise Tutoring
Practice difficult conversations	Role Play
Understand multiple perspectives and trade-offs	Debating
Build confidence through repetition	Formative Assessment Generator
Synthesise expert advice from different roles	The Expert Panel

5.10 Combining Techniques

You can also combine techniques into a professional workflow:

Student uses Reverse Prompting to scope out all requirements for a new policy
Student uses Pros and Cons to evaluate three different approaches
Student drafts their chosen policy
Student uses Role Play to practice explaining it to a skeptical stakeholder
Student uses Debating to analyse potential criticisms from different perspectives

This sequence takes the student through analysis, decision-making, communication, and critical reflection: a bounded rehearsal, not evidence of a complete professional capability.

5.11 Choose and test a pattern

Allow 12 minutes. Select a pattern and identify the learner action it supports. Write the source/criteria check and an independent follow-up. You can do this on paper using the pilot pack’s fixed responses.

For example, reverse prompting may help scope a review, but the learner must decide which question is relevant and why. A role-play can rehearse wording, but its emotional response is not a grade. A generated quiz needs checked answers before students rely on it.

Use the data gate for every live prompt and retain a non-AI route. The prompt library collects bounded versions of these patterns.

# Seven Essential Prompt Techniques for Business Teaching > The right technique is not the cleverest one. It is the one that matches the kind of thinking you want your students to do. ## Why These Seven? These seven prompting patterns offer different ways to organise practice. Their educational value depends on the task, evidence, feedback and learner activity; the list is not a claim that each pattern has been validated across every discipline. 1. **Reverse Prompting**: teaches comprehensive scoping and requirement gathering 2. **Pros and Cons**: builds analytical decision-making skills 3. **Stepwise Tutoring**: reinforces process adherence and documentation 4. **Role Play**: develops communication and interpersonal skills 5. **Debating**: strengthens strategic thinking and ethical reasoning 6. **Formative Assessment Generator**: supplies draft practice questions and provisional feedback 7. **The Expert Panel**: develops multi-perspective analysis and synthesis skills Each pattern needs adaptation to the discipline, supplied records and intended capability. The examples below use different disciplines to show the range, so adapt the prompts for your own field. These seven are adapted from *Conversation, Not Delegation*, which covers the same approaches for a general audience. If you want the underlying rationale for why each technique works, see the companion book. --- ## Technique 1: Reverse Prompting (for Scoping) Instead of solving a problem immediately, the AI asks the student questions to help scope out all requirements, considerations, and potential issues. This mirrors professional work where defining the problem is half the battle. **Example (HR: policy design):** ``` I need to draft a new "Flexible Work Arrangement" policy for a 500-employee company with both office-based and remote staff. Your task: Ask me a series of questions to help me clarify all the requirements, considerations, and potential pitfalls for this policy. Ask one question at a time. Wait for my response before asking the next. Continue until you've helped me think through at least 10 different aspects. Begin with your first question. ``` The AI asks questions the student must think through before responding: fairness, logistics, legal compliance, technology needs, cultural impact. The questions may broaden their scope, but cannot guarantee completeness. Compare the resulting scope with the supplied task criteria. **Adapt for your discipline:** Replace the policy topic with whatever scoping challenge fits: a market entry strategy, an audit plan, a supply chain redesign, an IT implementation. The technique is the same: AI questions, student thinks. **Teaching tip:** Use as a pre-writing exercise. Ask for a short evidence index and the subsequent draft. Assess the reasons for consequential decisions; a complete private chat history is not required. --- ## Technique 2: Pros and Cons (for Decision Making) The AI systematically analyses multiple approaches to a problem, evaluating each against specific criteria. Students must then critically evaluate the analysis rather than just accept it. **Example (Supply Chain: strategy evaluation):** ``` Compare different supply chain strategies for a mid-sized manufacturer: make-to-stock, make-to-order, and mass customisation. For each strategy, provide: 1. A brief description of how it works 2. Three key advantages 3. Three key disadvantages Evaluate each specifically in terms of: - Cost efficiency - Responsiveness to demand changes - Inventory risk Conclude with a recommendation for a company facing unpredictable seasonal demand and justify your choice. ``` **The follow-up task:** Check one consequential advantage and one disadvantage against supplied evidence. Identify a genuine omission if there is one. Students may retain the recommendation when it is justified; do not require a different strategy merely to demonstrate opposition. **Teaching tip:** Project the AI's analysis on screen. Divide students into groups, each arguing for a different strategy. They use the AI's framework but add their own reasoning. --- ## Technique 3: Stepwise Tutoring (for Process) The AI walks through a complex process one step at a time, pausing after each step until the student signals readiness. Use a checked teaching procedure and inspect each learner attempt. **Example (Accounting: audit process):** ``` I am learning how to conduct an audit engagement from planning through to the audit opinion. Walk me through the entire process. For each step, tell me: 1. What action to take 2. What to document 3. What professional standard or legal requirement applies at this stage Use the supplied teaching procedure and criteria. Do not invent standards or legal requirements. After each step, ask me to apply it to the supplied case and justify my decision. STOP while I answer. Identify missing information. Do not provide the entire process at once. Begin with Step 1. ``` Typing “next” would not demonstrate understanding. Inspect the student's application and justification against checked teaching criteria before proceeding. A generated explanation is not a view into the model's internal reasoning. **Teaching tip:** Have students repeat the exercise with complications introduced at each step: "What if the client refuses to provide documentation at Step 3?" Check any generated guidance against the supplied procedure before relying on it. --- ## Technique 4: Role Play (for Communication Skills) The AI adopts a specific persona and engages in a realistic conversation. This is the "flight simulator" concept: students practice difficult professional interactions where the AI responds dynamically. **Example (Management: difficult conversation):** ``` Play Alex in the fictional P1 practice pack. Use only P1-D1 to P1-D5. Open with the concern about dependencies in P1-D4. I am the junior adviser, not the decision-maker. Do not invent private history, policy or evidence. Say when an answer is not supplied. Stop immediately on STOP or after six exchanges. Do not require disclosure, agreement or a final ruling. I will respond to your opening statement. ``` The learner practises wording, evidence checks and a bounded next step. Text role-play does not establish competence in every aspect of a live workplace meeting. **Adapt for your discipline:** AI plays a demanding customer (marketing), a resistant employee (management), a skeptical auditor (accounting), a supplier announcing delays (supply chain), a stakeholder with unclear requirements (IT). **Teaching tip:** Ask for a relevant excerpt or the response to the fixed opening, with a short explanation of what the learner would retain or change. --- ## Technique 5: Debating (for Strategic and Ethical Analysis) The AI examines multiple perspectives on a contentious issue, argues both sides, and helps students see complexity and trade-offs. **Example (Economics: multi-stakeholder debate):** ``` Simulate a leadership debate on whether to implement a carbon tax. Create three personas who will each argue from their perspective: Persona 1: Environmental Economist Focus: climate impact, externalities, long-term economic benefit Persona 2: Industry Representative Focus: competitiveness, cost burden, employment impact Persona 3: Treasury Official Focus: revenue, fiscal policy, implementation feasibility Have each persona make their opening argument. Use only the supplied teaching sources. Label these as generated perspectives, not independent evidence. Do not invent empirical claims or citations. Then have them respond to each other through three rounds. After the debate, summarise the key tensions and what a compromise policy might look like. Begin the debate. ``` Students cannot just pick a side. They must understand legitimate competing perspectives and make a justified recommendation that acknowledges trade-offs. **Teaching tip:** This can support practice in analysing a case. Instead of memorising model answers, they practise analysing competing priorities. --- ## Technique 6: Formative Assessment Generator (for Practice) An educator can request draft questions, scenarios or quizzes for low-stakes practice. Review the question, intended answer and feedback before student use. **Example (Tourism & Hospitality: scenario practice):** ``` I'm studying for an exam on service recovery in hospitality management. Generate 10 short scenarios (2-3 sentences each) where I need to identify: 1. What went wrong in the service delivery 2. The appropriate recovery strategy 3. A possible impact on guest loyalty and what evidence would be needed to investigate it Use the supplied teaching criteria and answer notes. After each attempt, offer provisional feedback with its source. Flag ambiguity instead of inventing a definitive answer. Make the scenarios progressively more complex. Begin with Scenario 1. ``` Students can request further practice, subject to access and tool limits. Use checked questions and feedback, then an independent changed problem to examine learning. Repetition and a model's approval do not establish mastery. **Critical principle:** This is for practice, not graded assessment. Students need to struggle and make mistakes in low-stakes environments. --- ## Technique 7: The Expert Panel (for Multi-Perspective Analysis) Students consult multiple AI "experts" with different professional perspectives on the same problem, then synthesise the competing advice into a strategic recommendation. **Example (Information Systems: system implementation):** ``` Our company needs to decide on an ERP implementation approach. Create three expert personas: Expert 1: IT Architect Focus: technical feasibility, integration, security, scalability Expert 2: Change Management Consultant Focus: user adoption, training, organisational readiness, resistance Expert 3: CFO Focus: cost, ROI, risk, business case Treat these as generated viewpoints, not real experts or independent evidence. For each provide: 1. Their assessment of the key risks 2. Their recommended approach 3. Limitations of their recommendation 4. What data they would need to validate it Begin with Expert 1. ``` After receiving the perspectives, students compare the claims, check them against supplied records or independent sources, and justify a recommendation. Agreement among personas is not corroboration. **Teaching tip:** This technique combines well with Debating (have the experts debate each other) and Role Play (present your synthesis to a skeptical stakeholder). --- ## How to Choose Which Technique to Use | If you want students to... | Use this technique | |---|---| | Define a complex problem comprehensively | **Reverse Prompting** | | Evaluate competing options and justify a choice | **Pros and Cons** | | Follow a sensitive process correctly | **Stepwise Tutoring** | | Practice difficult conversations | **Role Play** | | Understand multiple perspectives and trade-offs | **Debating** | | Build confidence through repetition | **Formative Assessment Generator** | | Synthesise expert advice from different roles | **The Expert Panel** | ## Combining Techniques You can also combine techniques into a professional workflow: 1. Student uses **Reverse Prompting** to scope out all requirements for a new policy 2. Student uses **Pros and Cons** to evaluate three different approaches 3. Student drafts their chosen policy 4. Student uses **Role Play** to practice explaining it to a skeptical stakeholder 5. Student uses **Debating** to analyse potential criticisms from different perspectives This sequence takes the student through analysis, decision-making, communication, and critical reflection: a bounded rehearsal, not evidence of a complete professional capability. ## Choose and test a pattern **Allow 12 minutes.** Select a pattern and identify the learner action it supports. Write the source/criteria check and an independent follow-up. You can do this on paper using the pilot pack's fixed responses. For example, reverse prompting may help scope a review, but the learner must decide which question is relevant and why. A role-play can rehearse wording, but its emotional response is not a grade. A generated quiz needs checked answers before students rely on it. Use the [data gate](ethics-integrity.qmd) for every live prompt and retain a non-AI route. The [prompt library](../appendices/prompt-library.qmd) collects bounded versions of these patterns.