Article
How to Teach AI in Marketing Education without Sacrificing Critical Thinking
AI is changing how students approach their work, as tools like ChatGPT and NotebookLM can generate answers in seconds.
But marketing has never been about quick answers. It’s about judgment, trade-offs, and strategy. That leaves educators with a real challenge:
How do you teach students to use AI without letting it replace the thinking that matters most?
This article explores how Dr. Claudio Schapsis of Sacred Heart University used CAPTURE, his structured prompting framework, with the Stukent Social Media Marketing Simternship to help students use AI without harming critical thinking.
From casual prompting to strategic thinking
At the start of the course and experiment, most students approached AI the same way: quick questions, short phrases, and vague requests. As one student noted, "I used to ask quick, casual questions ... like how I would ask a friend. I always thought AI would just 'figure it out.'”
That approach led to generic outputs and, in turn, generic performance.
To address this challenge, Dr. Schapsis introduced students to the CAPTURE framework:
Context
Audience
Purpose
Tone
Unique Constraints
Role
Examples
The framework forced a cognitive shift, turning prompting into a deliberate skill. Students were required to slow down, stop looking for shortcuts, and define what they actually wanted before asking for it.
Students described the process as moving from chaotic to controlled. AI became more predictable and useful. The framework and application in the Simternship helped them realize that prompting is a skill and not as simple as it seems.
More importantly, they began thinking like marketers, not just prompt writers.
Read more on the CAPTURE framework
How AI enhances the Simternship learning experience
Once students started applying the framework inside the Social Media Marketing Simternship, the shift in their thinking became clear.
AI can generate ideas, but ideas without context fall flat. Inside the simulation, students worked with real constraints, real data, and real audiences.
The strongest outcomes came when students began using the simulation's built-in resources and feeding them into the AI. Instead of writing generic prompts, they drew directly from simulation materials, including the Reynolds Reports, Buhi Market Data, brand guidelines, and social audits.
This combination creates something powerful: a feedback loop between strategy, execution, and results.
The AI helped students with:
Persona matching: Students used buyer personas like Daypacker Tom or Back-to-School Mindy to guide AI-generated captions and headlines that actually fit their audience.
Strategic scheduling: AI helped analyze market data and best practices to build content calendars with clear timing and posting strategies for each platform.
High-volume ideation: Organic post ideas quickly scaled, with AI generating dozens of post ideas and giving students more options to evaluate and refine.
AI handled the volume, but students handled the thinking.
They had to go through the process of evaluating which ones were most appropriate, selecting what aligned with their strategy, and refining what didn’t.
That’s the same workflow modern marketing professionals use every day.
The most important lesson: AI can be wrong
Perhaps the most valuable lesson for educators is that AI was not a "cheat code" for a perfect score. In fact, several students stated that when they relied too heavily on AI, their performance dropped.
In one case, when a student asked AI for help determining the cause of poor results, it hallucinated reasons, blaming “audience fatigue,” when data showed the real problem was incorrect targeting. In another instance, a student requested step-by-step instructions from the AI and saw their lowest score of the semester.
These AI errors taught students that AI doesn’t understand strategy; it just follows instructions. They learned that AI isn't a substitute for their work or the field, but it can be a helpful thought partner with a clear strategy.
One student added, “AI can only follow what you tell it, not actually understand the brand ... using AI works best when you already have a clear strategy."
Creating AI-ready marketers
By the end of the simulation, students weren't just better at using ChatGPT; they were better marketers. Students left the course with a better understanding of AI, with takeaways such as:
AI speeds up ideation, not strategy
Better inputs lead to better outputs
AI is good for tone adjustments
Human judgment drives results
Or as one student advised: “Use AI to help you think, not to think for you.”
The CAPTURE framework gave students a clear way to turn generic inputs into thoughtful, strategic prompts, while the Stukent Simternship created the ideal environment to put that into practice. With layered data, evolving scenarios, and defined constraints, students had the complexity needed to truly test AI, not just use it.
One student reflected on how their approach changed: “… immediately after using the CAPTURE framework, I saw a difference in the response I was getting and the results of the simulation.” That shift wasn’t just about better prompts; it showed a deeper understanding of how to guide AI with purpose.
Together, the CAPTURE framework and the simulation created a space where students could experiment with AI, see it fail, and learn why.
FAQs
How should students use AI in marketing simulations?
Students should use AI for brainstorming and analysis, feeding and validating outputs with real data from the simulation.
What is the CAPTURE framework?
It’s a prompting method that enhances AI output by setting context, audience, purpose, tone, unique constraints, role, and examples. It aims to guide students to use AI as a thought partner, not a replacement for thinking.
Why do students struggle with AI?
Because they find it hard to create good prompts. AI can sound correct even when wrong. Without strategy and validation, outputs can mislead students and hinder learning.
Watch the webinar on teaching strategic prompting
What’s the best way to teach AI in marketing?
Combine a structured framework with hands-on simulation environments so students connect inputs, outputs, and results.
