Photo by: Supatman via Canva
If you’ve ever opened an AI tool, typed in a request, and thought, “Hmm… that’s not quite what I needed,” you’re not alone. Many of us have been there. The good news? It’s likely the prompt, not AI.
A well‑crafted prompt is the difference between a generic output and results you can actually use in your design work. Whether you’re generating a task analysis, writing a case study, or creating an image for a slide deck, the quality of your prompt determines the quality of your output.
The secret is simple: every strong prompt has three essential steps. Let’s break them down.
STEP 1: Prepare the Framework
Before you type a single word into your AI tool, pause. The biggest mistake designers make is jumping straight into the request without giving AI any context. Framework is all about building the context AI needs to produce accurate and instructionally sound output.
Start by identifying the chunk of content you want help with. Maybe it’s a task, topic, or situation. The more specific you are, the more focused the output will be.
Then add the essentials:
- What you’re trying to create (a script, quiz question, image, method design, etc.)
- Any background or training strategy details
- Who your learners are, and what they already know
- The phase of instruction you’re designing for
- Any examples you want the AI to model
Think of this step as briefing AI the same way you would for a colleague who’s jumping into your project. A little clarity goes a long way.
STEP 2: Specify the Parameters
Once the framework is set, you’re ready to shape the output. In this step, you tell AI how you want the content to look, sound, and feel.
Format
Give your AI tool a clear structure to follow. You might want:
- A short paragraph.
- Bullet points.
- A step‑by‑step sequence.
- A script or dialogue.
- A specific word count or length.
Style
Style is all about the writing itself. You can guide AI by asking for:
- Simple or more complex sentences.
- Short, punchy phrasing or longer, flowing explanations.
Tone
Tone sets the mood. You might want something:
- Conversational
- Formal
- Serious
- Enthusiastic
Words to use or avoid
A quick list can make a big difference. For example:
- Use action verbs, keywords, positive language.
- Avoid jargon, vague words like “good,” subjective terms like “appealing,” or unnecessary metaphors.
Additional details
Include names, places, dates, stats—anything specific you want included. These small details help AI produce accurate content.
STEP 3: Write the Command
Now you’re ready for the final step: the actual instruction. In this step, you tell AI exactly what to do.
Always start with an action verb. Try verbs like:
- Generate
- Create
- Design
- Write
- Develop
For example: “Generate a short case study scenario using the details above.” Or “Design five quiz questions based on the task analysis.”
Why These Steps Work
When you combine these three parts—framework, parameters, and a clear command—you get content that’s relevant, structured, and almost ready to use. The final piece is to ensure you’ve fully validated the output.
AI isn’t here to replace instructional designers, but it can help us work smarter, faster, and with more creativity. A little intention behind your prompts can go a long way in streamlining your workflow.
Additional Resource: The Secret to Better AI Outputs [Video]
If you want to go deeper and build these skills with hands‑on practice, join us for the Enhancing Instructional Design with AI workshop. Check our website for the next available date and come learn with us.