I am an instructional designer working on [PROJECT CONTEXT] for [LEARNERS]. Create a copy-and-paste prompt pack for ChatGPT, Claude, or Gemini that helps me complete the recurring work below. Include one ready-to-use prompt each for: 1. A discovery-question set for a subject-matter expert, separating must-know performance from background information. 2. Measurable learning objectives using observable verbs, conditions, and criteria. 3. A course outline that maps each objective to instructional activity, practice, feedback, and assessment.
4. A scenario-based practice activity based on [WORKPLACE SITUATION]. 5. A knowledge check with plausible distractors based on common mistakes, plus rationales for every answer. 6. A review of existing learning content for accuracy risks, cognitive overload, accessibility, and objective alignment. 7. A launch evaluation plan using [SUCCESS MEASURES], with leading and lagging indicators. For each prompt, include only the inputs I need to replace in square brackets, clear instructions to the AI, and a specified output format. Make the prompts usable in a corporate learning setting, not academic essay prompts. Require the AI to distinguish facts supplied by the SME from assumptions; never fill an evidence gap with invented policy, process, or compliance information. Use [DESIGN CONSTRAINTS] to account for delivery format, seat time, tools, approval requirements, and accessibility needs. Build in appropriate checks: objectives must be assessable, every assessed skill must have practice first, and the assessment must measure job performance where feasible rather than recall alone. Before presenting the pack, check that all seven prompts are materially different and that each names a concrete deliverable. Ask up to 3 clarifying questions only if a required input is missing.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [PROJECT CONTEXT] | Describe the business problem, topic, organization, and desired on-the-job change. | New-hire training for a regional bank's customer-service team on handling address-change requests while preventing account fraud. |
| [LEARNERS] | Describe the learners' roles, prior knowledge, location, language, and typical work environment. | 120 call-center representatives with 0–6 months of tenure, working in English on monitored phone calls. |
| [WORKPLACE SITUATION] | Provide a realistic job situation in which learners must make a decision or perform a task. | A caller sounds rushed, passes basic verification, then asks to change an address and send a replacement debit card immediately. |
| [SUCCESS MEASURES] | List the operational, quality, behavior, or learner measures that define success. | Fewer missed fraud escalations, correct case documentation, quality-assurance score improvement, and learner confidence after 30 days. |
| [DESIGN CONSTRAINTS] | State the delivery channel, available time, systems, accessibility needs, and review constraints. | A 35-minute self-paced module in Rise, screen-reader support, legal review, and no use of real customer records. |
How to use
- Replace the five bracketed inputs with the scope notes, learner profile, and constraints you already have.
- Paste the full prompt into your AI tool and save the resulting prompts in your team’s standard template location.
- Check that each objective can be observed on the job and that every assessment target has preceding practice.
- Follow up with: “Rewrite prompt 4 for a branching scenario with two credible incorrect choices and feedback tied to each choice.”
Variations
SME interview guide
Use this before course design when the task and source material are still unclear.
Create a 45-minute SME interview guide for [TOPIC] with [SME ROLE]. Organize questions into business outcome, high-stakes decisions, expert-versus-novice differences, error consequences, source-of-truth materials, and examples of successful performance. Mark which answers require a document or policy citation. End with a capture table for task steps, decisions, common errors, and evidence. Do not infer procedures that the SME has not confirmed. Ask up to 3 questions only if required inputs are missing.
Assessment blueprint
Use this when objectives are approved and you need an assessment before building content.
Build an assessment blueprint for [LEARNING OBJECTIVES] for [LEARNERS]. Create a table with objective, performance evidence, assessment method, scenario or item type, scoring rule, common misconception, and required practice before testing. Favor job-realistic decisions and artifacts over isolated recall where possible. Identify objectives that cannot be validly assessed in [DELIVERY FORMAT]. Do not write policy details not present in the inputs. Ask up to 3 questions only if required information is missing.
Storyboard review
Use this to quality-check a draft module before SME or stakeholder review.
Audit this storyboard: [STORYBOARD]. Evaluate it against [LEARNING OBJECTIVES] and [ACCESSIBILITY REQUIREMENTS]. Return a prioritized issue log with slide or section, issue, why it matters, and exact revision. Check for unsupported claims, content that does not serve an objective, practice after assessment, vague feedback, excessive on-screen text, inaccessible interactions, and untested assumptions. Preserve accurate content; do not invent facts to repair gaps. Ask up to 3 questions only if a required input is missing.
Tips
- Start discovery with the job task and the consequence of an error; a content inventory is not a performance analysis.
- Write an assessment blueprint before drafting slides so the course does not drift into untested background knowledge.
- For multiple-choice items, make distractors reflect real novice errors, not obviously silly alternatives.
- Treat SME statements as inputs to validate against approved sources, especially for legal, safety, and process content.
FAQ
Can AI create a complete instructional design document?
It can draft a useful outline, storyboard, and assessment plan, but a designer still needs to validate task accuracy, evidence, and organizational constraints with the right owners.
How do I stop AI from producing vague objectives?
Give it the job context and require observable behavior, conditions, and performance criteria. Reject verbs such as understand, know, and become familiar with unless they are tied to observable evidence.
Should I give an AI tool internal course material?
Follow your organization’s approved AI and data-handling policy. Remove customer, employee, and confidential information if the tool is not approved for it.