AI Industry Agents: Simulating Conflicting Voices in Sport Marketing

When I asked sport industry professionals earlier this year what they wished graduates knew, the message wasn’t about technical tools or platforms. It was about reality:

• You don’t only answer to one person.

• You hear feedback from multiple directions, often conflicting.

• You rarely get to run the campaign you’d design in a classroom.

• You work in an ecosystem, not a silo — clubs, agencies, sponsors, and governing bodies all have a say.

That’s the truth students often miss. University can feel streamlined — one lecturer, one rubric, one grade. Industry isn’t like that.

This is why I am experimenting with a suite of AI industry agents into my sport marketing course. Not to provide tidy answers, but to simulate the messy, layered conversations that happen in real campaigns.

How the course is structured

This course is built around four case briefs, each grounded in real NZ sport organisations: HYROX, Moana Pasifika, Golf New Zealand, and ARL/Warriors. Each mini-brief includes the organisation’s background, strategic priorities, and what they expect from a marketing campaign.

Students select one brief and choose a single priority area to focus on. From there, they:

• Identify a target market.

• Conduct market research (secondary sources, participation data, industry reports).

• Develop a marketing strategy that aligns with the brief.

• Design a campaign concept, including commercialisation opportunities.

This is the first assessment — a strategic marketing campaign built from the ground up.

The second assessment raises the stakes. Students get one week to respond to a disruption: something in the environment has changed (e.g., policy, participation numbers, sponsor exit, media coverage). They must write a memo that shows how they would adapt their campaign in response.

This scaffolding matters. Students don’t just pitch once and walk away; they experience how campaigns evolve when reality intervenes.

Why an agency model?

I framed the whole project around a fictitious agency — Clubhouse — rather than positioning students as working inside a club. Students are acting as a Junior, and can pick one of their ‘client briefs’ to work from.

That was deliberate. Sport organisations don’t always have the capacity to run campaigns in-house. They brief agencies because agencies bring strategy, commercialisation, and activation expertise.

By situating students “inside” Clubhouse, I give them the chance to feel that dual role:

• Accountable to their boss (Agency Director).

• Accountable to the client (sport organisation).

• Accountable to commercial partners who might activate or fund their idea.

This three-way accountability is a truer reflection of the sport ecosystem.

Meet the Agents

To mirror the messy ecosystem of sport marketing, I designed a suite of six AI industry agents. Each brings a different perspective that students must navigate:

• Sophie – Agency Director (Clubhouse)
  The boss. Sophie gives sharp, strategic feedback, pushing students to make their campaigns clear, distinctive, and professionally polished. She asks the hard questions you’d expect in a real agency pitch.

• Levi – Commercial Partnerships Manager (Clubhouse)
  The commercial voice. Levi stresses brand fit, ROI, and activation. He pushes students to think beyond “logo on jersey,” consider measurable outcomes, and explore sustainable ways campaigns create value for both clubs and partners.

• Georgia – Pathways & Engagement Manager (ARL/Warriors)
  The dual-hat client. Georgia balances ARL’s grassroots priorities (participation, retention, community) with the Warriors’ elite goals (fan engagement, brand visibility). She often sounds conflicted, because that’s the reality: Rugby League has to compete with other sports in Auckland, and campaigns must work for both grassroots and elite.

• Jayden – Youth Participation Manager (HYROX)
  The growth client. Jayden focuses on how HYROX can introduce its youth fitness programme into Auckland schools. He pushes students to show how campaigns create confidence, clear participation pathways, and feel authentic to NZ youth culture — not just competitive fitness.

• Pele – Community Impact Manager (Moana Pasifika Trust)
  The community client. Pele makes sure campaigns reflect Pasifika identity and values, strengthen belonging in South and West Auckland, and create opportunities for youth and women. Authenticity is non-negotiable; flashy campaigns that don’t align with Moana Pasifika’s kaupapa won’t fly.

• Amelia – Inclusion & Pathways Manager (Golf New Zealand)
  The inclusion client. Amelia represents Golf NZ’s push to make the game more accessible and inclusive. She challenges students to tackle perceptions of elitism, open up genuine pathways for rangatahi, women, and Māori, and show how golf can connect to lifestyle and wellbeing, not just tradition.

Together, these six voices create triangulated feedback: strategy from the boss, ROI from the commercial manager, and authenticity from the client — each one testing campaigns in a different way.

What is Cogniti?

These agents weren’t built with an off-the-shelf chatbot. They were deliberately engineered using Cogniti, a generative AI platform designed for educators to create structured AI agents without needing developer skills.

Cogniti has been taken up by universities and education providers across Aotearoa, Australia, and Singapore since its launch in 2023, and is now used to build everything from tutors and mentors to role-play simulations. Its big advantage is that it lets educators design custom agents that stay in character, anchored to context, and focused on learning outcomes.

For me, the benefits are clear:

• Lock in roles: every agent begins with “You are acting as…” so students always know whether they’re talking to Sophie the Agency Director, Levi the Commercial Manager, or a client rep.

• Control the style: short, conversational, 4–6 sentence responses that sound like real meetings — not essays.

• Anchor to real briefs: HYROX agents will always push on youth fitness, Moana Pasifika will always push on authenticity, Golf NZ will always push on inclusion.

• Tone consistency: Sophie stays sharp, Levi pragmatic, client reps protective. They don’t drift into generic advice.

• Closed system: Cogniti only works off what I feed it. It can’t search the internet or hallucinate irrelevant answers. That means I control the direction and keep feedback focused on the case briefs.

• Back-end analytics: I can see how students are interacting — what trends, tones, and sticking points are emerging. This helps me understand how they’re navigating industry-style conversations.

This matters because it prevents drift. A normal chatbot might hand out generic marketing tips or go rouge with random information. Cogniti keeps each agent in character, sounding like a real professional voice students would encounter in the industry and, crucially, never gives away the answer.

How students can use the agents

The agents won’t appear on day one. I hold them back until after Week 6, once students have developed their own target markets, dug into market research, and sketched out campaign ideas. That’s deliberate — just like in the real world, you wouldn’t walk into a client or agency meeting with nothing in hand.

From Week 6 onwards, students can start “working with” the agents when they’re ready to test their thinking:

• Visit Sophie when they want to sharpen their strategic “why” or refine the professionalism of their pitch.

• Bounce an idea off Levi to stress-test the commercial angle and think about ROI.

• Check in with a client rep when they’re not sure if their idea genuinely fits the organisation’s priorities.

That timing matters. By making students do the groundwork first, the conversations with Sophie, Levi, and the client reps feel earned and the feedback lands more like a real industry exchange.

Why they don’t give answers

In real life, yes, sometimes clients or bosses do give answers. But in the classroom, if these agents solved things for students, it would short-circuit the learning.

That’s why I set a hard rule: no solutions, only sharper questions. Each response follows the same breadcrumb model:

• Affirm something promising.

• Link it back to the brief.

• Raise a concern or gap.

• Ask a question that forces students to justify, refine, or rethink.

This deliberate friction matters. The agents aren’t designed to make life easier; they’re designed to create the same productive tension students will face in their first marketing job.

Case anchors in practice

When students do start using the agents, the feedback will sound like this:

• HYROX rep:
  “Yeah, schools run events all the time — sports days, cross-country, PE fitness units. What actually makes your idea stand out as HYROX, not just another school tournament? And once the kids have had a go, how are you keeping them connected to HYROX beyond that one day?”

• Moana Pasifika rep:
  “I can see where you’re going, but would our South and West Auckland families actually see themselves in this campaign? Or does it risk looking like it was designed from the outside? That’s the test for us.”

• Golf NZ rep:
  “Running another clinic is fine, but clinics don’t shift the dial. How does your idea actually change the way people see golf — especially women, rangatahi, and Māori — rather than reinforcing the same old stereotypes?”

• ARL/Warriors rep:
  “From the Warriors’ side, yep, we get visibility and brand exposure. But with my ARL hat on, I’d ask — does this do anything to keep kids playing league at their local club? Because if it doesn’t, we’ve just created hype without fixing the bigger problem.”

No answers. Just sharper questions.

Why this matters

By the end of the course, students won’t just have built campaigns. They’ll have:

• Learned to manage multiple stakeholder perspectives.

• Practised defending and refining ideas under conflicting pressures.

• Confronted the commercial reality of ROI, not just creativity.

• Grappled with cultural authenticity and identity in sport marketing.

• Experienced what it feels like to work inside an agency, navigating clients and partners.

This is what industry told me students need to learn: that marketing is ecosystem work. It’s never just you and your idea. It’s you, your boss, your client, your commercial partners and sometimes they don’t agree.

That’s why I built these agents. Not to give answers, but to create the conditions where students learn to think harder, defend stronger, and navigate the messy, real-world dynamics of sport marketing.

Looking ahead

We haven’t rolled the agents out yet, they’ll be introduced next week, once students have done the groundwork on target markets, research, and campaign design. I want them to pitch with substance, not blindly.

And at the end of semester, I’ll be sharing a follow-up blog on how the experiment worked: how students used the agents, what surprised me, and what it taught them about the complexity of sport marketing.

Because for me, this isn’t about replacing teaching with AI. It’s about making classrooms feel just a little more like the real world students are about to enter.