Why this is hard to get right
Imagine you're a management professor at a regional university. You've just agreed to swap course sections with a colleague mid-semester, which means you're inheriting three case studies you didn't select and have never taught.
The first case lands on your desk on a Tuesday. Class is Thursday morning. You have two other courses to prep, a department meeting, and a stack of ungraded papers.
You open the instructor's manual that came with the case — if one exists at all. It gives you a synopsis and maybe five generic discussion questions. There's no guidance on sequencing, no warnings about where students typically get stuck, no suggested time splits, and no debrief framework. You're on your own.
So you do what most instructors do: you over-prepare the content side and under-prepare the facilitation side. You know the case cold by Thursday morning, but when the room goes quiet after your opening question, you're not sure whether to push harder or pivot. The discussion meanders. Students hit the obvious surface-level points. The deeper strategic tension in the case — the thing that makes it worth teaching — never fully surfaces.
This is the gap a well-constructed teaching note fills. Not just "here are some questions," but a sequenced facilitation guide that moves students from recall to analysis to genuine judgment. One that tells you where to expect confusion, how to open the discussion with enough tension to spark debate, and how to close with takeaways that stick.
The challenge is that writing a teaching note at this level takes 2-3 hours when done properly. Most instructors don't have that time — especially for a case they didn't choose. AI can compress that work dramatically, but only if you give it enough context to produce something useful rather than a generic placeholder.
That's exactly the problem a structured, context-rich prompt solves.
Common mistakes to avoid
Omitting the Session Length
Without a time constraint, the AI generates more discussion questions than any class could realistically cover. Always specify your session length so the AI can build a pacing guide that reflects how much ground you can actually cover.
Skipping the Student Profile
A teaching note for first-year undergraduates needs very different scaffolding than one for experienced MBAs or executive learners. Omitting the audience level produces questions that are either too elementary or too abstract for your actual students.
Asking for Questions Without Specifying Sequence
Requesting 'discussion questions' without specifying cognitive sequencing produces a flat list. Tell the AI to progress from comprehension to analysis to evaluation — or reference Bloom's Taxonomy — to get a properly scaffolded discussion arc.
Forgetting to Specify Output Sections
If you don't list the sections you need, the AI guesses. You may get a synopsis and questions but miss the debrief framework, time allocation, or misconception warnings that make a teaching note genuinely useful in the classroom.
Using a Generic Case Description Instead of Naming the Case
Describing a case as 'a company facing an ethical dilemma' forces the AI to invent details. Naming the actual case — or providing a 2-3 sentence summary — grounds the teaching note in real context and produces far more relevant discussion questions.
The transformation
Write a teaching note for a business case study about a company making a difficult decision.
**Act as an experienced business school instructor** with expertise in case-based pedagogy. Create a complete instructor teaching note for a case study about **Patagonia's 2022 ownership restructuring decision**, designed for an **MBA-level Business Ethics and Strategy course** (25-30 students, 80-minute class session). **Include the following sections:** 1. Case synopsis (150 words max) 2. Learning objectives (3-4 measurable outcomes using Bloom's Taxonomy action verbs) 3. Discussion questions (6 questions, sequenced from comprehension to evaluation) 4. Suggested time allocation per discussion phase 5. Common student misconceptions to address proactively 6. A "cold call" opening question to launch discussion 7. Key takeaways for the board debrief **Tone:** Collegial, practical, and analytical. Write for an instructor who may be teaching this case for the first time.
Why this works
Role Precision
Assigning the AI the role of 'an experienced business school instructor with expertise in case-based pedagogy' activates a specific knowledge domain. The AI draws on pedagogical conventions — discussion sequencing, cold calls, debriefs — rather than producing generic educational content.
Named Context
Specifying the actual case, course name, and student level eliminates ambiguity. The AI doesn't have to invent a scenario — it can anchor every section of the teaching note to real strategic, ethical, or analytical content from that specific case.
Taxonomic Anchoring
Referencing Bloom's Taxonomy for learning objectives signals that outcomes should be measurable and cognitively differentiated. This prevents vague objectives like 'students will understand the case' and produces action-verb outcomes like 'students will evaluate stakeholder trade-offs.'
Structured Output Mandate
Listing the exact seven sections with formatting instructions means the AI produces a usable document, not prose that needs to be reformatted. Numbered sections, word limits, and section titles all become output structure when specified in the prompt.
Audience Framing
The closing instruction — 'write for an instructor teaching this case for the first time' — shifts the register from academic to practical. It tells the AI to explain facilitation decisions, not just list questions, making the note genuinely useful for someone without prior case experience.
The framework behind the prompt
Case-based pedagogy traces its modern form to Harvard Law School in the 1870s and was later adapted for business education at Harvard Business School in the early 20th century. The method rests on a core constructivist principle: learners build deeper, more transferable knowledge by actively grappling with realistic, ambiguous problems rather than receiving information passively.
The instructor's role in case discussion is not to lecture but to facilitate a structured inquiry that moves students through increasingly sophisticated levels of reasoning. This is where Bloom's Taxonomy becomes a practical tool. Originally published by Benjamin Bloom in 1956 and revised in 2001, the taxonomy identifies six cognitive levels — remember, understand, apply, analyze, evaluate, create — each demanding progressively more complex thinking. A well-sequenced teaching note ensures discussion questions ascend this ladder rather than stalling at the comprehension level.
A second key framework is discussion mapping, a facilitation technique where instructors pre-plan how student responses connect to each other and to the case's central tension. Teaching notes that include board plans and anticipated misconceptions operationalize this technique, giving instructors the structural support to facilitate rather than just manage conversation.
The AI prompt strategies on this page draw directly on these principles — using role assignment to invoke pedagogical expertise, cognitive-level labeling to enforce question sequencing, and structured output sections to mirror the components of a professional teaching note.
Prompt variations
Act as an experienced law school clinical instructor trained in the Socratic and problem-based learning methods.
Create a complete instructor teaching note for a 1L Contracts course case involving Hadley v. Baxendale (consequential damages rule), designed for a 65-minute class of 60 students using cold-call format.
Include:
- Legal issue synopsis (100 words)
- Three measurable learning objectives (use action verbs: identify, apply, distinguish)
- Eight cold-call questions sequenced from issue spotting to policy analysis
- Anticipated student errors in applying the foreseeability test
- Hypo variations to extend the discussion if time allows
- Board plan outline for capturing key doctrine
Tone: Rigorous and Socratic. Assume students have read the case but not the secondary material.
Act as a senior corporate learning facilitator with experience designing leadership development workshops.
Create a facilitation guide for a 60-minute case workshop using an internal scenario about a mid-level manager who mishandled a team conflict during a product launch, designed for 15-20 high-potential employees in a leadership pipeline program.
Include:
- Workshop framing statement (read aloud by facilitator, 60 seconds)
- Learning objectives (2-3 outcomes tied to our leadership competency framework: communication, accountability, decision-making)
- Small group discussion questions (4 questions for 20-minute breakout)
- Large group debrief questions (3 questions)
- Common defensive responses to anticipate and how to redirect them
- Key messages to reinforce in closing
Tone: Direct, psychologically safe, and growth-oriented.
Act as an experienced AP Social Studies teacher skilled in document-based and case-based instruction.
Create a teaching note for a 50-minute class discussion of a case study on the 1954 Brown v. Board of Education decision and its implementation challenges, designed for 30 AP U.S. History students (Grade 11) who have completed the assigned reading.
Include:
- Case background summary (100 words, student-accessible language)
- Learning objectives aligned to AP U.S. History Historical Thinking Skills (causation, continuity and change, argumentation)
- Five discussion questions progressing from factual recall to historical argumentation
- One primary source excerpt to anchor the discussion
- Exit ticket question to assess comprehension
- Common misconceptions about Brown's immediate impact to address directly
Tone: Engaging, academically rigorous, and appropriate for advanced secondary learners.
When to use this prompt
Business School Faculty
Professors creating teaching notes for Harvard-style case discussions need structured facilitation guides that align to MBA learning outcomes and handle sophisticated student pushback.
Corporate Learning & Development Teams
L&D professionals adapting real company scenarios into internal case studies for leadership development programs need teaching notes that fit 60-90 minute workshop formats.
Online Course Creators
Educators building asynchronous courses on platforms like Teachable or Kajabi need written facilitation notes to guide learners through self-directed case analysis without an instructor present.
Law and Medical School Instructors
Professional school faculty using problem-based learning need teaching notes that scaffold clinical or legal reasoning from basic comprehension through complex judgment calls.
High School AP and IB Teachers
Advanced secondary educators using case-based methods in economics, history, or social sciences need age-appropriate discussion scaffolds tied to curriculum standards.
Pro tips
- 1
Specify the exact session length so the AI can allocate realistic time blocks to each discussion phase — a 50-minute class needs very different pacing than a 90-minute seminar.
- 2
Name the cognitive level you want students to reach by end of class (e.g., 'evaluate trade-offs' vs. 'identify stakeholders') so the discussion questions build toward that endpoint deliberately.
- 3
Include your student profile — class size, prior course experience, and whether they've read the case in advance — because the AI will adjust the cold-call question and misconception warnings accordingly.
- 4
Add any constraints you're working around, such as 'students have not studied Porter's Five Forces yet' or 'this is the first case discussion of the semester,' so the teaching note scaffolds appropriately rather than assuming background knowledge.
Most teaching notes fail because discussion questions all target the same cognitive level — usually basic comprehension or surface-level analysis. Bloom's Taxonomy gives you a six-level framework to build a properly sequenced arc:
Level 1 — Remember: "What decision did Patagonia's founder make?" Use this to open and confirm students have read the case. Keep it brief.
Level 2 — Understand: "Why did Yvon Chouinard structure the transition the way he did?" This checks comprehension of the case logic before analysis begins.
Level 3 — Apply: "Using stakeholder theory, map the parties affected by this decision." Students apply a framework from the course to a new situation.
Level 4 — Analyze: "What tensions exist between Patagonia's environmental mission and its obligations as a business?" This is where most discussions should spend the majority of time.
Level 5 — Evaluate: "Was this the right decision? What criteria are you using to judge it?" Forces students to make and defend a judgment, not just describe the situation.
Level 6 — Create: "If you were advising a different company facing a similar values conflict, what framework would you recommend?" Use this for advanced courses or as an optional extension.
When you specify this sequence in your prompt, the AI generates questions that build on each other rather than repeating the same analytical move six times.
A board plan is a pre-designed layout of what you intend to write on the whiteboard or display slide during discussion. Including one in your teaching note request dramatically improves classroom coherence — students can follow the discussion arc visually, and you avoid the common trap of capturing student comments randomly without structure.
To request a board plan in your prompt, add this instruction:
"Include a suggested board plan showing how to visually organize key concepts, stakeholder positions, and decision trade-offs as they emerge during discussion."
A good AI-generated board plan will typically include:
- Column headers that reflect the case's core tension (e.g., Mission vs. Growth)
- Row labels for stakeholder groups or decision criteria
- A 'parking lot' section for interesting student comments that don't fit the main framework but shouldn't be dismissed
- A synthesis box at the bottom where you'll capture the class's collective takeaway
For online or hybrid courses, a board plan translates directly into a shared Google Slide template that students can annotate in real time — which makes the facilitation note doubly useful.
Not all case discussions follow the same format. Your prompt should reflect the specific structure you're using:
Harvard Business School-Style (Full Case, 80+ minutes) Focus on a single central decision with full financial, operational, and strategic context. Your prompt should request 6-8 discussion questions, a detailed protagonist analysis, and an exhibit-specific discussion section if the case includes data exhibits.
Mini-Case or Vignette (20-30 minutes) Strip the teaching note to essentials: one sharp opening question, three focused discussion questions, and a two-minute debrief close. Specify 'mini-case format' explicitly in your prompt.
Multi-Stage Case (Split across 2 sessions) Request teaching notes for each stage separately. Session 1 should end on a decision point without resolution; Session 2 should open by revisiting student predictions before revealing what actually happened. Ask the AI to include a 'session bridge' — a short framing statement that reconnects the two sessions.
Decision-Forcing Case (No right answer revealed) Ask the AI to write the teaching note without a recommended solution section, and instead request a framework for helping students evaluate the quality of their own reasoning rather than comparing to an 'answer key.'
When not to use this prompt
This prompt pattern is not the right fit when you're working with a highly confidential internal case that you cannot describe to an AI system, or when your institution has specific copyright restrictions on case material that prevent you from inputting case details into a third-party tool.
It's also less effective for purely skills-based training — like a software tutorial or a step-by-step technical procedure — where a teaching note is less relevant than a structured exercise or worked example. In those cases, use a task-based instructional design prompt instead.
Troubleshooting
AI generates discussion questions that are all the same cognitive level
Add explicit sequencing instructions to your prompt: 'Generate questions that progress through Bloom's Taxonomy from comprehension (Level 2) to evaluation (Level 5). Label each question with its cognitive level.' This forces the AI to distribute questions across the full range rather than defaulting to analysis-level questions throughout.
Teaching note is too generic and doesn't reference the actual case content
Paste a 3-5 sentence case summary directly into the prompt, or name specific exhibits, characters, and decision points you want the discussion to address. The more specific your case context, the more specific the AI's questions and facilitation notes will be. Generic case descriptions produce generic teaching notes.
Time allocations are unrealistic for the session length
Specify your session length at the top of the prompt AND ask the AI to justify its time splits. Add: 'For each discussion phase, explain why you allocated that amount of time given the cognitive demand of the questions.' This self-checking instruction catches allocations that would leave no time for debrief or student questions.
How to measure success
A successful AI-generated teaching note should pass four checks before you use it in class. First, each learning objective should contain a specific Bloom's Taxonomy action verb and describe an observable behavior. Second, discussion questions should clearly escalate in cognitive demand — you shouldn't be able to swap question 2 and question 6 without disrupting the arc. Third, the time allocation should add up to your session length with at least 10 minutes reserved for debrief. Fourth, at least one anticipated misconception should feel genuinely surprising — if they're all obvious, the AI was too conservative and you need to add more case-specific context.
Now try it on something of your own
Reading about the framework is one thing. Watching it sharpen your own prompt is another — takes 90 seconds, no signup.
a case study teaching note for your course
Try one of these
Frequently asked questions
Yes — and it often works even better. Instead of naming a published case, paste a 3-5 sentence summary of your custom scenario directly into the prompt. The AI will use that summary as its source material and generate discussion questions, learning objectives, and facilitation notes grounded in your specific context.
Reference a cognitive framework explicitly in your prompt. Bloom's Taxonomy is the most widely recognized: ask for questions that progress from 'remember and understand' through 'analyze, evaluate, and create.' You can also describe your students directly — for example, 'students who have completed one prior strategy course' — so the AI calibrates question complexity accordingly.
Specify the shorter session length and ask for a 'compressed facilitation guide' with no more than three discussion questions and a single key takeaway. The AI will prioritize depth over breadth, which is exactly what a short session requires.
Absolutely. Add a line specifying 'asynchronous online delivery' and note that there is no live facilitator. The AI will reframe discussion questions as written reflection prompts, replace cold-call guidance with independent analysis tasks, and shift facilitation tips toward self-directed learner engagement.
Specify that each objective must begin with a Bloom's Taxonomy action verb and include an observable behavior. For example: 'Evaluate the ethical trade-offs in Patagonia's decision using stakeholder analysis' is measurable. 'Understand the case' is not. Telling the AI this distinction upfront produces objectives you can actually assess.