"He’s Like You, But in the Screen”: What Jack the AI Agent Taught My Class

In large sport management cohorts, we often talk about the challenge of scale, how to bring industry relevance into the classroom without relying solely on guest talks or one-off events. In 2025, I trialled a new approach using an AI industry agent called Jack, designed to simulate Auckland FC’s Marketing Manager. Jack wasn’t just a gimmick, he helped prompt real change in the way my students thought, wrote, and worked together. And I’ve got the numbers, quotes, and reflections to back it up.

If you haven’t read the original post about Jack’s design, context, and setup — start here: Using AI Industry Agents to Enhance Learning in Sport Management.

A Quick Recap: Who’s Jack?

Jack is a custom AI agent hosted in Cogniti, designed to act as a marketing expert for Auckland FC. His job? Help students test and refine ideas in Assessment 3: a group marketing pitch responding to a real challenge facing the club.

He’s not just a chatbot; he’s built to respond like a real stakeholder, pushing student thinking, asking strategic questions, and challenging assumptions. He knew the club's values, strategic objectives, target audiences, and recent initiatives. He was also updated regularly with news articles, player signings, and developments to stay relevant.

How I Introduced Jack

Jack was introduced in Week 9 of the course, after students had already completed Assessments 1 and 2. By this stage, they had researched Auckland FC’s operating environment, identified strategic challenges, and developed stakeholder maps. What they needed next was support in refining their creative solution — and that’s where Jack came in.

Here’s how I framed it to students:

Jack is your AI marketing manager from Auckland FC. He’s here to help you shape a smarter, sharper, more realistic idea for your final pitch.

• He acts like a real industry contact — but one everyone in our class can actually talk to at the same time.

• He’ll challenge you to think deeper, especially around audiences, sponsors, and platforms.

• He’s designed specifically for this course, using what we know about Auckland FC and marketing trends.

Students were encouraged to use Jack for:

• Sense-checking their idea

• Suggesting content formats

• Identifying logical gaps or audience mismatches

• Linking ideas to strategic insights or sponsor value

• Testing tone, brand fit, or creative execution

I set homework for the students to 'test' Jack. They were asked to:

• Chat to Jack at least once

• Try two different types of prompts

• Bring one useful insight or change to class the following week

We also reminded them to be clear, professional, and specific and to take screenshots or notes for reference.

In Week 10, we spent time in our groups planning out their assessments (with no technology)

1. Clarify the challenge: What is the issue you're solving, and why now?

2. Explore who it affects: Beyond fans, who else is impacted?

3. Design a strategic initiative: What could Auckland FC actually launch?

   
   

   Once they had a rough plan, they could then ask Jack for feedback.

Suggested Jack prompts included:

• "Here’s our idea for Auckland FC — can you help us improve it?"

• "How can we make this more strategic for stakeholders?"

• "What are some common reasons fans don’t come back?"

• "Does this sound like something that could actually work?"

  
  

I walked the room, sat with groups, and co-created alongside them, acting as a real-time guide and sounding board while Jack gave structured support that we worked wioth and developed together.

What shifted using Jack

• Students consistently started with vague or literal ideas such as “engage with youth,” “grow social media,” “build loyalty.” Jack redirected them with plain, strategic questions like “what’s the real opportunity here?” or “how might fans respond to that?”

• A key pattern emerged: students shifted from describing a problem to exploring its strategic and emotional dimensions. It wasn’t just “get more fans” — it became “connect with parents who are choosing a sport for their child” or “make loyalty meaningful when success isn’t guaranteed.”

• Three out of four students refined or reimagined their idea during their conversation with Jack — a sign that the right questions can spark real strategic thinking.

• Students wrote pages of notes. I walked past desks filled with handwritten prompts, rewritten challenges, and arrows connecting ideas. One said, “this actually helped me figure out what we’re really trying to do.”

• There was a buzz , students stayed late, bounced ideas in groups, and used Jack’s questions as brainstorming fuel. “He’s like you, but in the screen,” one said to me. “It’s like you’re with us when you’re not.”

• I even heard a group shout out in class, "oh KIA ORA [thank you] JACK!!!" as they struck gold in a conversation

  
  

  Survey Data

To gauge the impact more formally, I ran a short survey with one cohort after students completed their conversations with Jack.

• 87% agreed or strongly agreed that Jack helped them develop their idea further.

• 83% said Jack helped them think more critically.

• 100% said they would use an AI industry agent again in another class

  
  

  Student Open Quotes

• “Jack didn’t just answer our question but also made us expand our knowledge and keep on going with our thinking.”

• “Jack fleshed out ideas and wanted us to do the same when communicating with him. + The helpful little ‘Great ideas, that would work’ made the interaction impressionable.”

• “We had grand ambitions without realistic restrictions (imagination running wild), while Jack retained a real-world grounded approach with possible consequences.”

• “Jack would get back to me with many stakeholders then at the end would create a focus question for me, like ‘why are these stakeholders important? What affect does gender equality have on them?’”

• “Jack helped to refine my idea as before talking to jack my idea was unclear and very broad which I wouldn’t have known without Jack.”

• “Jack helped us think more critically and gave us more ideas to develop and focus on.”

• “I had no idea what idea I wanted to go with before the lecture, then talking to him during the lecture helped me come up with a bunch of different ideas that I could pitch to my group.”

• “It was sort of like having a proper conversation with someone and really helped me come up with ideas for the assessment to make sure I fit the guidelines too.”

• “Very informative, such a cool idea, makes it easier than asking the teacher can just do it yourself.”

• “Yes, Jack didn’t give us the answers but he helped propped our ideas to think more creatively.”

•  “Jack suggested multiple ways on how to keep our LinkedIn post precise, formal and to the point. Jack also prompted us to develop our idea further so that it met the rubric requirements of explaining the challenge and how it impacts AFC.”

Key Insights from the Chat Data

I reviewed over 150 individual conversations with Jack. Most started with short, uncertain prompts and then grew. It wasn’t just a tool students used to “check” an idea. It was a space to clarify, question, test, and sometimes scrap what they’d planned entirely. Here’s what stood out.

Students treated Jack like a person and a partner

Across the data, students regularly said “Hi,” “Thanks,” and even “Have a great day.” Some opened with full explanations, like they were talking to a new colleague. Others asked, “Can I run something past you?” or “Does this sound okay?” “Hey Jack, we’re thinking of doing something for Pacific fans… not sure if it’s a dumb idea though.” This conversational tone mattered. It showed trust. Students weren’t interrogating a tool, they were co-creating with a partner.

Strategic blind spots showed up fast

The most common early prompts were vague:

• “We want to do a giveaway.”

• “Something for students.”

• “We just want to build loyalty.”

Jack’s questions revealed how surface-level many of these starting points were. Repeated gaps included:

• Not knowing why their idea mattered now

• Oversimplified audiences (e.g. “young people” as a catch-all)

• Lack of rationale behind content formats (e.g. “TikTok because it’s popular”)

• Weak sponsor alignment (“we could put a logo on it”)

But when Jack asked questions like:

• “What does loyalty mean in a sport with inconsistent success?”

• “Why would Go Media care about this?”

• “Who’s excluded from your idea?”

— the shift was often immediate.

Students layered complexity across the chat

Conversations often followed a trajectory like this:

Vague idea - Clarified challenge - Reframed audience - Adjusted initiative -Linked to strategy - Chose execution format - tested tone

“Okay, that’s actually a much better idea. We might start again.”

Jack wasn’t offering a pre-set path, he was prompting layers. This process helped move students from “marketing as promotion” to “marketing as purpose.”

Tone and wording mattered more than we thought

Students weren’t just asking “what should we do?” , they were testing language, brand voice, and phrasing. Some used Jack to trial:

• Campaign titles

• Slogans

• Instagram post copy

• LinkedIn captions

They asked: “Does this sound strategic?”, “Would a sponsor back this?”, and “Is this cringe?”

“It feels like a joke — how could we reword it so it sounds more real?”

Jack’s feedback loop helped them match tone with audience , a difficult but critical part of sport marketing.

There were real ‘a-ha’ moments

You could see it in the logs. Students would say “we’re not sure” or “we just want to build awareness” and then, five messages later, they were mapping loyalty over a season, designing a TikTok series for injured academy players, or talking about intergenerational belonging in Pacific families.

💬 “Wait… I actually love that. We didn’t even think about the cultural side.”

These were moments of conceptual shift. Not just improving an idea, changing the thinking behind it.

Jack gave students language for thinking

Students often knew what they wanted, but not how to express it. Jack helped label ideas (“this is a values-led campaign,” “you’re targeting time-poor fans,” “this is retention, not reach”), giving students the vocabulary to move from concept to communication.This language support also helped with writing their LinkedIn-style post, even if the full drafts Jack gave weren’t quite usable.

They didn’t just listen — they pushed back

Jack was challenged in multiple chats:

“I don’t think that would work for our group ,do you have another idea?” “I don’t want to do a social media campaign, what else could we try?”

This matters. Students weren’t passive. They used Jack as a sounding board, questioned him, and redirected him. The best conversations weren’t just helpful, they were collaborative.

Students gave Jack context and trusted him with it

Many chats included references to prior assessments, lecture content, and personal ideas: “Our A2 was about loyalty, so we’re thinking of building on that…” “We did a SWOT and identified the sponsor challenge…” They didn’t just ask generic questions, they embedded Jack in their process. That level of transparency is hard to get even from real guest speakers.

Some conversations broke the limits of what I expected

In one standout thread, a group designed a digital tool for academy players to help transition to senior squad life, with sponsor backing. It didn’t come from Jack. It came from a student's idea that Jack kept pushing: “Why does this matter?” “What’s the actual problem?” “Who would fund it?”

By the end, the group had a full 3-tier sponsorship model, community impact layer, and TikTok strategy to match. And it started with: “We want to make a TikTok about young players.”

Challenges & Lessons

Jack wasn’t perfect. There were some clear friction points:

• There was a moment mid-marking when I panicked: two groups had submitted almost identical ideas. For a split second, I thought Jack was giving students templated, repeatable answers, but after digging deeper, I realised it was just a teammate re-sharing. Still, it flagged something important: some students had used Jack’s words verbatim in their submissions. This wasn’t widespread, but it highlighted a tension. I had to clarify, to students and in Jack’s own backend prompt, that he was a thinking tool, not an answer bank.

• Sport changes week to week. Auckland FC had new player signings, community updates, and ongoing shifts in narrative. To keep Jack relevant, I had to update his context regularly — feeding him news articles, club press releases, and stakeholder commentary. When I didn’t, his responses felt static. Lesson learned: if your agent is embedded in a live case, you have to treat it like a living resource too.

• Jack tended to give long, structured responses. While thoughtful, they overwhelmed some students who were just looking for a nudge. Even after editing his default reply structure, the conversational flow was still a bit ‘essay-like’. Next time, I’ll break replies into more manageable chunks — short, scaffolded questions rather than mini-lectures.

• One of Jack’s roles was to help students write a LinkedIn-style post pitching their group idea. The output? Polished but robotic. Students said it helped clarify tone and structure, but it lacked personality. What they really needed were prompts to develop their own voice, not a script. I’ll shift this to prompt-only support next round.

• Embedded inside Canvas, Jack’s conversations couldn’t be saved or revisited. This made it hard for students to refer back to helpful exchanges, especially when moving from idea to final write-up. It broke the feedback loop. In future iterations, we’ll ensure conversation history is exportable or savable, even if it’s just a screenshot function built into the platform.

• Not every student used Jack. Some preferred to brainstorm as a group without tech, others weren’t sure what to ask, didnt come to class or didn’t feel confident using AI tools and a few simply forgot. These moments matter. They’re not signs of failure, they’re reminders that AI literacy and digital fluency must be actively scaffolded, not assumed. Also, AI should not be compulsory as part of an assessment, but a choice.

  
  

Each challenge became a design opportunity, pushing me to rethink not just how Jack talks, but what we ask him to do.

Final Thoughts

Despite the bumps, Jack helped because he was:

• Grounded in context. He knew Auckland FC; their values, audiences, and strategy.

• Conversational. He asked, rather than told.

• Reflective. He didn’t hand over ideas, he teased them out.

• Scaffolded for challenge. Jack didn’t just support — he pushed students to think harder.

That’s why students walked away with sharper thinking, stronger ideas, and a better understanding of what sport marketing strategy actually looks like. This was reflected in their final marks, with a high grade average for creative thinking.

What’s Next: A smarter, more responsive agent ecosystem

This pilot is just the beginning. Here’s where I’m thinking about heading next:

• A suite of agents with distinct voices: a community officer, digital strategist, commercial lead, or cultural advisor.

• Built-in prompts that encourage idea development, not templated answers.

• Better UI integration so students can revisit and revise past chats.

• Optional agent tones , mentor, challenger, reflective partner, so students can match the feedback style that works best for their learning.

• AI-delivered Wild Cards. Instead of static disruptions, I’m trialling agents that can drop project-specific challenges, like funding cuts or media controversies, tailored to students’ ideas, to simulate real-world pivots.

Ultimately, I want to move beyond one agent and toward a dynamic, diverse classroom of digital mentors. Because you can’t scale yourself, but you can scale thinking.

This wasn’t about using AI because it’s trendy. It was about giving students a sounding board they could actually access. Not once. Not just if they happened to be free when a guest came in. But consistently, whenever they needed to test or grow their thinking.