Talent in financial services: an interview with Laureen Chivere

1. You have built and transformed learning functions at both Naspers and ASML. When starting from a fragmented or immature environment, what do you believe leaders need to get right first?

“The most important thing is understanding the value chain of the business. How do its functions actually contribute? Where are people and capability the real constraint on delivering the strategy? That’s where the opportunity sits.

Fragmented doesn’t mean broken. In an immature or fragmented environment, it usually just means nobody has asked why each piece grew the way it did. Understand that first. Connect it to the strategy. Look for what you already have in common. That’s what lets you build a shared vision: one answer to why you all exist, and what you’re here, together, to do.

When I joined Prosus in 2016, that’s what finding the common ground looked like in practice. We already had excellent in-person leadership programs, including partnerships with Harvard and Stanford. I wanted to build on that, not replace it. So we built an online learning platform that gave every employee across the group real access to development. Founders could focus on growth, while we built leadership, succession, product, tech, data science and culture capability underneath them, at scale. The shared vision went further than content, too. We connected people across companies and countries who had genuinely never been in a room together. That became a marketplace for sharing talent and knowledge.

At ASML, the same principle ran in the other direction. There was already a great depth of learning expertise across the business. It had simply grown up in pockets. Each pocket was strong on its own, but there was no shared vision connecting them, and no easy way to build on each other’s work. I wanted to connect it into something bigger: one shared identity and architecture, even though delivery stayed federated across the business, not one centralized machine. What came out of that work was one ASML Academy, built together rather than imposed on anyone. It cut the time it took someone to reach real competence by 40 percent. That’s what let us onboard more than 10,000 new employees in a single year, and more in other years after.”

2. At ASML, you brought fragmented learning teams together into a more centralized global academy. How did you decide what needed global consistency and where local teams needed the freedom to adapt to business and cultural context?

“We approached it together, and we grounded it in data. We looked at what the business and our people were actually asking for. We looked honestly at everything we were already offering. That picture showed us exactly where the opportunity was. We could simplify a long list of learning providers. And teams solving similar problems in different parts of the business could learn from each other, instead of working in parallel without knowing it.

From that, we built an operating model together. A center of expertise and shared services brought the overlapping pieces into something more efficient. Safety training and core technical certifications stayed global, because the risk of getting those wrong was the same everywhere. How a manager coached a team, or onboarded new employees, stayed with the business, because that’s where it actually worked best. Our learning steerco shaped that architecture with us, and our Academy leaders. The thinking came from across the whole learning team, not just the people at the top of it.

We put Talent and Learning leaders in every country, so local relevance was built into the model from the start, not bolted on afterwards. We didn’t stop at our own walls either. We asked other companies what had worked for them, and more usefully, what hadn’t. That way we weren’t paying to learn lessons someone else already had.

We piloted it. We came back with what was working and what wasn’t. We kept adjusting, until what we delivered good results. We kept iterating and adjusting how we delivered it, ensuring the ROI from the business confirmed with data that it was working well.”

3. Learning is often seen as a support function rather than a business priority. How have you built the case for learning and capability as a strategic investment, particularly when resources, budgets and attention are under pressure?

“One thing I say often: compare two companies in the same industry, with exactly the same number of employees. The real differentiator is almost never the product. It’s the culture and the skills inside the business – that’s the real competitive advantage. The return on investing in people is exponential, not linear. ASML is the clearest proof of that I’ve seen. Learning there wasn’t just formal training. It was mentoring, coaching and communities of practice, woven into the culture everywhere. Not a line sitting apart from the business.

Making the case stick with a CFO meant dropping learning’s own language. What engages our leaders in those conversations is speed, cost and risk. That’s the currency the rest of the business already uses. We built dashboards that made our work visible every quarter. We used real ROI measures to see what was actually working, and removed what wasn’t, down to the bottom line, to behaviour, to the skills that genuinely transferred back into the job and into business and sector results. That discipline, together with finance and procurement leaders across the group, let us rationalize the learning portfolio. We cut duplication. We cut what didn’t add value. We aligned what was left to the capability gaps that actually mattered to the strategy. The result was a 43% reduction in enterprise-wide learning spend, while we actually extended reach and served double the workforce from when we started. It also cost something real. Walking away from a few programs people were genuinely attached to, and those conversations were harder than the one with the CFO but we were all happy that these were the right decisions for the business.”

4. When redesigning a learning operating model at scale, what are the most common mistakes organizations make, and what does effective change leadership look like in practice?

“Most of the mistakes I’ve seen come from the same instinct. Leaders look at what worked somewhere else, and assume they can plug it in a new environment and context. The missing step is always the same. You have to understand how this particular organization actually works, how it’s structured, and what its operating model really needs, before deciding what structure would serve it best.

The easiest version of that mistake, and the one I see L&D professionals make most often, is moving straight from decentralized to fully centralized. It’s the obvious, tidy answer. It photographs well in a steering committee deck. It isn’t always the right one. At ASML, we ended up with a federated model. I hadn’t built one before, and wouldn’t have chosen it out of habit. But it fit the way the rest of the business was structured because it aligned with the business operating model.

Change leadership, in practice, is much slower. It involves far more stakeholder management, and managing far more competing priorities. It means bringing the people who built the old model into the room designing the new one, instead of informing them afterwards. It means agreeing, early and out loud, which few things the business genuinely can’t bend on, so people aren’t left guessing for months. And it means deciding together which old ways of working to let go of, agreeing what replaces them, and holding each other accountable to make it stick.

A launch proves nothing on its own. The real test comes a year later: is the business genuinely being served better than it was before?”

5. AI is changing both the skills organizations need and the way people can learn. How should HR leaders approach AI-enabled learning without allowing the technology to become the strategy?

“AI is a capability requirement and a delivery mechanism. Those are two different things, and they keep getting collapsed into one. The strategy question hasn’t changed. Where do we want this business to go? What skills does that require? What’s the real gap between where our people are and where they need to be? Only once that’s answered should you ask where AI can accelerate the route there. Specifically, where it should assist, where it should augment, and where it can genuinely automate. Assisting means it helps someone work faster, and the person stays responsible. Augmenting means it improves the judgement call itself. Automating means it runs on its own, inside guardrails someone deliberately set. At Prosus, we weren’t chasing maximum automation. Just the right balance for that particular piece of work.

The moment your roadmap gets described by which tool you bought, you’ve already gone wrong. Deloitte’s own research backs that up: 93% of AI investment still goes to the technology, and only 7% goes to the people meant to use it. Adoption was never a technology problem for us. It was a behavior-change problem, and we treated it as one.

HR has to lead that conversation in the open, with people, so they can make informed choices about their own next steps and the direction of their career. When AI genuinely takes over part of someone’s role, they deserve a straight answer about what that means for their job. Not silence. One of our executive assistants put it best. AI didn’t take her job. It gave her back her time. The real work was helping her find where she added value now. That’s a job-architecture question as much as a training one. It means partnering with business leaders to redesign the operating model and the roles themselves.

Some of that thinking has to come from outside your own walls. IKEA’s retraining of call-center staff, once those roles became redundant, is the example I point to most. It created value for the business and a new purpose for the people. It wasn’t a trade-off between the two. None of it works as a single rollout either. I treat adoption as something to prototype, then adjust and keep adjusting, because the pace of change is faster than anything we’ve seen before. The organizations that keep pace with that are the ones still standing when it slows down.”

6. What will distinguish organizations that successfully build AI fluency across their workforce from those that run a series of pilots but struggle to create sustained adoption?

“Pilots are designed to prove the technology works. Sustained adoption is designing for something harder. It has to prove the technology changes what people do every day, when nobody’s watching.

The research backs that distinction up. MIT’s 2025 study on the state of AI in business found that 95% of generative AI pilots fail to deliver any measurable return. Not because the models don’t work. Because of the gap between the tools and the organization actually using it.

That’s the tell. Are employees still using it on an ordinary Tuesday? Not just a workshop.

The honest metric is never completion. It’s whether people are still opening the agent unprompted, weeks after the novelty wears off, because they’ve genuinely started to depend on it. Programs that stay pilots produce testimonials that use the word “exciting.” The ones that stick produce something quieter: real, unglamorous reliance. The clearest predictor of which one you get is leadership. BCG’s research on AI at work found regular use running at 85% among leaders, but only 51% among frontline employees. That gap has barely moved. If leaders don’t visibly use something themselves, the frontline won’t either. That’s the fastest way to keep a pilot stuck as a pilot.”

7. Having led across Africa, Europe and global organizations, what have you learnt about leading through different cultures, and how should HR leaders balance global standards with local relevance?

“I believe diverse teams make better decisions, and that shows up in results, especially during change and transformation. McKinsey’s research backs that up: companies with the most diverse leadership teams, by gender and by background, are 39% more likely to outperform the least diverse ones. I’ve seen that play out across the many markets I’ve worked in, and most clearly in the transformations I’ve led. What I’ve learned leading across them is that what actually changes, culture to culture, is rarely the substance of an idea. It’s how directly you can give feedback. It’s how much consensus a decision needs before it’s real. It’s what “urgent” actually means in a given market. That’s why I default to the thinking of the whole group, not just my own read of a room.

That belief shapes how I balance global and local. I pair global direction with real local autonomy, because that combination creates responsiveness. It’s the ability to serve local needs better, and more easily, than any single playbook could. A few things genuinely have to be the same everywhere. Safety is the clearest example. Everything else is decided locally, including the right to tell me a global idea won’t work in their market. Even something as practical as translation shows this in action. Done properly, it tells a local team they’ve been seen, and that their culture is valued. Get that balance right, and you get something rare: global solutions that actually stick, with local customisation exactly where it matters.”

8. If a CHRO asked you where to focus first in order to build a more future-ready organization, what would you advise them to do in the next 12 months?

“I’d start with CHROs genuinely getting on top of AI themselves. Understand how it actually works. Don’t just sponsor it from a distance. Sit as a peer with the CTO or Head of AI, so you’re both looking at the same picture: where AI can disrupt the business, and where it can accelerate the strategy, in the short term and the long term.

From there, do an honest capability inventory against where the business is actually going. Not the workforce plan that’s three years old and hasn’t been adapted. Most learning strategies I’ve seen solve yesterday’s gap very well. I’d also push every CHRO to look at the new job architectures AI is creating around the world right now and work out which model actually fits their own organization. And I’d align with senior leaders early on how they see AI’s role in their function and organization, so that the message to employees during Ai transformation stays consistent across every function. One voice, not five different ones.

Then I’d pick one capability, one real risk to the business/function, and go deep there, instead of spreading effort across everything.

A CHRO who can point to a few capability areas the business measurably changed in twelve months carries more weight than one holding a list of twenty initiatives that each moved a little.

The organizations that move fastest in year one are the ones brave enough to let go of what no longer serves them. That’s how an organization builds real agility.”

9. Measuring the impact of capability building is an area where many learning functions struggle. Beyond participation, completion rates and learner feedback, how can HR and learning leaders assess whether their investment is genuinely changing behaviour and improving business performance?

“Completion and satisfaction are the easiest things to measure, and the easiest to report upward. That’s exactly why they’re overused. They tell you people showed up. They don’t tell you anything changed.

The discipline is agreeing the real business metric with the business owner before the program is built, not after. Cycle time, error rate, time-to-hire, whatever it is.

If you can’t name the metric before you build the program, you’ll struggle to find it afterwards.

Then look for behavior-level evidence. What does a manager actually see changing? Not only what the learner says about themselves.

I remember one leadership program I built with my team that scored beautifully on every survey. A year later, it hadn’t moved the capability gap it was funded to close. That was uncomfortable to report. It was also the most useful thing we ever did. We stopped the program, and built one that actually showed a real change in behavior.

We trained most of the learning organization at ASML on the Philips ROI methodology, and built the infrastructure so people could measure it themselves, for every program, regardless of size. Our team was later recognized for that work with the Best Worldwide Implementation of ROI, by the ROI Institute in 2024.”