Monica Paoluzzi

When you're driving AI transformation at scale, what are you actually trying to achieve? And what is the right approach to ensure the AI transformation proceeds at pace, and goes in the right direction? In our marketing team at JET, we've gone almost entirely bottom-up so far. And it's worked — the energy is real, the use cases are self-generated, the agents are solving actual problems. But I keep coming back to two questions. Is a bottom-up approach fast enough? In large, complex organisations, organic adoption can take months to reach the workflows that matter most. At some point, does leadership need to set hard targets — not just inspire? Bottom-up builds genuine capability and ownership, but top-down drives pace. I'll leave it to you if that means setting specific OKRs, working on a permanent hiring freeze, or assigning your team 'mission impossible' tasks that would necessarily require a heavy use of AI to succeed. And the second question: how to balance psychological safety with velocity? One of the biggest blockers of AI upskilling is the unspoken fear: "What does this mean for me?" Pooja Adiga made a point at Savant Retail Congress in Amsterdam last month about framing the narrative around amplification, not replacement. Expand what people can do. Don't just shrink what they need to do. I don't disagree with this. But here's the uncomfortable tension: if we're too protective of psychological safety, do we risk losing the momentum altogether? Because while we carefully (and rightly) manage the human side of this transition, something very different is happening elsewhere. Leaders across companies in the Prosus group spent a week embedded with AI leaders in China. Timo Boldt wrote about it in a way that made me stop and think. The framing there isn't about adoption curves or change management. It's about redesigning organisations from scratch around systems that don't sleep. Finance teams are already comparing token cost to headcount, not as a thought experiment, but as a live budgeting exercise. I think the message for us is clear: If you don't see how AI will fundamentally reinvent your business model, someone else will. So, here's where I've landed, for now. We need to separate two distinct objectives — because they require very different approaches. 1. Adoption across the organisation: start bottom-up, build genuine capability and ownership, then layer in top-down pressure gently to accelerate. This is for everyone. 2. Reinventing the model: this needs a strong top-down mandate from day one, and probably a small, focused group of people with the brief and the air cover to think from scratch. You can't crowdsource this one. Confusing the two is where most organisations will likely get stuck. Is your organisation moving fast enough — or are you just moving carefully? (Thanks Pooja and Timo, for the conversations and writing that sharpened my thinking here. And Tim Davis for the debate and the coaching!)

My Marketing Ops team set a target for the first quarter of 2026: get 50% of our marketing organisation actively using AI by end of quarter, from close to zero. 🎯 We hit 52%. And that's just the headline. Here's what the team actually built in one quarter: — 433 AI agents, created by marketers solving their own workflow problems — 114 validated use cases across 8 departments — 3 learning products designed and launched (from knowledge hub to hands-on hackathons) — 86 colleagues trained through workshops — 13 AI Champions mobilised across the org The number I find most telling? Content creation was the #1 use case in this wave (63% of agents built). Not because we told people to start there — because that's where the daily struggle was. Teams identified their own bottlenecks and built their own solutions. That's what good adoption looks like. Not compliance. Not a rollout deck. People solving real problems, at pace. 🚀 Proud of Maja Mitrović, Ieva Misiulaitytė, Franziska Walder, and Rebecca Williams for making this happen — and curious about what comes next. Because here's the honest question I'm now sitting with: we've built the foundation. We have adoption numbers, agent counts, and use cases logged. But is this enough to foster organisational change at the speed required to stay competitive? More on that in my next post 👇

The most expensive thing in a big organisation isn't inefficiency. It's the conversations that don't happen. When I moved to the Netherlands, one of the first things I noticed was how differently feedback worked here. In some cultures, a challenging comment arrives so ‘sugar coated’ that it's barely recognisable by the time it lands. Here, most of the time, people just say it: directly, kindly, and without the theatre. At first it surprised me. Then it felt like a relief and I started to understand why it matters at scale. This week at JET, we're rolling out a practice called "Address the Elephant in the Room" — a shared cue anyone can use to surface what's being avoided. A small elephant prop in the meeting room. A 🐘 in Slack. A low-pressure signal that says: I think there's something we need to talk about. In complex, multi-market organisations, the things that slow you down most aren't the hard problems. They're the obvious ones that somehow never make it into the room. The decision that keeps circling. The risk everyone's spotted but nobody has flagged. An obvious failure being treated with polite silence. When those things stay unsaid, you lose weeks. Sometimes months. Naming the elephant isn't about blame. It's not about conflict. It's about being honest while staying kind — so the team can focus on what will actually move things forward. I think my Dutch friends figured this out a long time ago. But building it into our company’s operating model is refreshing — and a genuine competitive advantage. What's the elephant in your room this week? 🐘

🚀 Excited to support our team as we grow! If you’re a strategic thinker with a passion for operational excellence in marketing, take a look at these two fantastic opportunities in Global Marketing Operations. Two pivotal roles, one great team. 👇