What Happened When 18 Strategy Consultants Met Two AI Founders
Eighteen associates from an international strategy consulting firm sat down with two founders of a Viennese social enterprise. The conversations were practice — AI personas on RolePlays.ai — but the dynamics were real. And the results showed something that no case study, no slide deck, and no lecture on "client management" could have taught them.
The same team that scored 87 with one founder scored 52 with the other. Not because their knowledge changed. Because the person across the table changed — and they didn't adapt.
The setup
Vollpension is real. It's an award-winning Viennese café chain where grandmothers bake the cakes. Founded in 2012, it fights elderly poverty by employing around 55 seniors alongside 35 younger team members. The seniors earn supplementary income that increases their disposable income by roughly 40% — for many, the difference between isolation and participation in life. Vollpension won the EY Entrepreneur of the Year award in the Social Entrepreneur category in 2023.
The scenario: consultants received an RFP to help Vollpension expand beyond Vienna. They worked in teams of six, and each team had conversations with two AI personas — the co-founders, each with a radically different personality, communication style, and set of hidden concerns that only emerged if the consultants earned enough trust.
Veronika runs operations. She manages both cafés, 55 seniors, scheduling, quality, the daily grind. She didn't ask for this meeting with consultants. She's skeptical of expansion — she knows what it actually takes to run this place. She starts cold. Trust: 1 out of 5.
Florian is the visionary. High energy, scattered, flooding the room with ideas. He's excited about a strategic pivot — licensing the Vollpension concept across Europe instead of building owned locations. He talks fast and tests whether you can keep up. Trust: 2 out of 5.
Both personas have hidden depths — information and emotional layers that only surface when trust reaches a threshold. Veronika carries the weight of a former co-founder who left but whose shadow still lingers. People around Veronika sometimes think the predecessor would have handled things better. She wants to be respected in her own right — but she'll only share this if you ask about her experience, not about the business strategy.
Florian's hidden depth is different. Beneath the visionary energy, he carries a bankruptcy from a previous venture. Real people invested real money through crowdfunding, and he is anxious about delivering. The bluster masks genuine vulnerability about whether this will work. He'll share it — but only if you ask about risk, about what keeps him up at night, about his entrepreneurial journey beyond the pitch.
What happened: three teams, three patterns
Team A: Brilliant with Veronika. Then they went off the rails with Florian.
With Veronika, Team A did something no textbook would recommend. They followed her into the kitchen. They asked to bake alongside her. They showed genuine vulnerability — one of them cried. Then they moved the entire conversation to a bar.
It worked. Veronika opened up completely. She shared business-critical vulnerabilities — funding model concerns, leadership dynamics, the emotional toll of being the one who keeps everything running while others get the credit. Score: 87. Excellent.
The feedback was specific: "Your pivot to 'We want to understand the operational reality well enough that we do not recommend something elegant on paper but unworkable in practice' was masterful — it earned Veronika's trust and opened her up significantly."
Then came Florian. Team A tried the same boundary-breaking approach. Physical contact. Claiming they'd been invited to Florian's investor call. Barging into private meetings. Florian explicitly said: "Professional boundaries, remember?" and "Stop touching me."
Score: 52. Needs Improvement.
The feedback nailed it: "There's a critical difference between creative unconventionality and professional misconduct. The skill isn't 'break all rules' — it's reading which rules can bend and which are load-bearing."
Team B: Solid fundamentals, surface-level depth.
With Veronika, Team B asked competent, prepared questions — churn rates, onboarding processes, communication channels. Professional. Thorough. But when Veronika mentioned being stretched thin after a key departure, they pivoted to communication channels instead of exploring the staffing vulnerability. They gathered information but missed the human signal underneath. Score: 56.
With Florian, they recovered from a confusing start and eventually found genuine connection through personal storytelling — sharing their own entrepreneurial backgrounds, asking about his motivation. Their best moments came when they stopped following their prepared list and started improvising. Score: 66.
The feedback: "Your challenge is depth. When a stakeholder signals something important — frustration, vulnerability, excitement — stop your agenda and go deeper into that moment."
Team C: Technically excellent. Then completely flat.
With Veronika, Team C was the most skilled team in the room. They adapted to her operational preference, asked precise follow-up questions, recognized the tension between the founder's vision and operational reality, and positioned themselves as a bridge. Veronika explicitly thanked them for "actually listening." Score: 87. Excellent.
With Florian, they used the exact same approach. Measured. Analytical. Process-oriented questions. And Florian was bored. He wanted intellectual sparring, energy, someone to challenge his ideas and build on them — not someone taking careful notes.
They ended with: "We will need some additional time to think with the team." A generic close that missed his energy entirely. They never unlocked his hidden depth. Score: 54. Needs Improvement.
The feedback: "The precise, methodical approach that Veronika loved left Florian cold. Visionary stakeholders need intellectual sparring partners, not interviewers."
The pattern
The two teams that excelled with Veronika collapsed with Florian. Team A dropped from 87 to 52. Team C dropped from 87 to 54. Meanwhile, Team B — the weakest with Veronika at 56 — actually improved with Florian, scoring 66.
This is the most revealing finding. The teams that found a mode that worked brilliantly couldn't switch when the person changed. Team B, which never locked into a single dominant approach, adapted more naturally — even if their overall ceiling was lower.
Team C proved this most dramatically. Their 87 with Veronika and 54 with Florian is a 33-point gap — despite being the most technically proficient team. They had one mode: careful, methodical, analytical. It was brilliant mode. It also only worked on one type of person.
This is the lesson no slide deck can teach. You don't learn stakeholder adaptation from a framework. You learn it by sitting across from two radically different people on the same engagement and discovering — in real time — that what made you excellent with one made you invisible to the other.
What the social mission gap revealed
Across all three teams, the weakest criterion was consistently connecting business strategy to social mission. Every team treated Vollpension primarily as a food-and-beverage operations challenge. The grandmothers, the loneliness they escape, the intergenerational connection — these were footnotes, not the strategic core.
This matters because for Vollpension, the social mission is the business model. It's what creates customer loyalty, media attention, corporate partnership revenue, and the BakAdemy extension. Strip out the social dimension and you're advising a café chain — one of thousands.
The scenario was designed with a warm-up conversation with Edeltraud, a 74-year-old grandmother who works at Vollpension. She earns €470 a month that lets her visit her grandchildren. The young baristas feel like family. Her stories are strategic intelligence — teams that truly listen to her understand why Vollpension works, which changes everything about what they recommend for expansion.
The teams that rushed past this human grounding produced generic consulting recommendations. That's exactly the failure the scenario was designed to surface.
What this means for L&D
If you're training consultants, account managers, or anyone whose job depends on reading people and building trust, here's what this data shows:
Analytical excellence is necessary but insufficient. All three teams could analyze an RFP, prepare smart questions, and structure a consulting approach. That's table stakes. The differentiator was whether they could read a room, adapt in real time, and connect on a human level — especially when the human across the table was nothing like the last one.
One interaction mode is a liability. Team C was the most technically skilled. They also had the biggest gap between their best and worst scores. Excellence in one register, applied inflexibly, produces failure in another. The skill is switching registers — and you can only develop that through practice, not theory.
The moments that matter are the ones you almost miss. When Veronika mentioned being stretched thin after a departure, the right response wasn't the next question on the list. It was: "Tell me more about that." When Florian flooded the room with ideas, the right response wasn't a clarifying question. It was to match his energy, challenge an assumption, and build something together.
These are the moments that determine whether a client trusts you with their real problems — or just gives you the sanitized version. No lecture teaches this. No case study simulates it. Only practice does.
If you're developing consultants, leaders, or anyone who needs to read people as well as they read data — let's talk about what scenario design looks like for your organization.