How Culture, Not Code, Is Powering Ogilvy’s AI Era

Workshops, drop-in clinics, peer-to-peer knowledge sharing and even a good old fashioned poster campaign. Artificial intelligence may represent the ultimate tech revolution, Ogilvy EMEA’s drive to encourage its people to get stuck in and start using AI has been surprisingly human – or perhaps it shouldn’t be all that surprising, given Ogilvy’s expertise in behaviour.

Over the past two years, since gaining access to WPP Open, the holding company’s AI platform, Ogilvy EMEA has been able to grow adoption from 20% to 80%. Overseeing this growth are Patou Nuytemans, CEO of Ogilvy EMEA and Dickon Laws, president of product strategy for WPP Open and EMEA AI transformation officer for Ogilvy.

For Patou, it certainly isn’t her first rodeo when it comes to adland tech disruptions, not least because she spent a good 12 years as EMEA chief digital officer and has a career with Ogilvy stretching back to 1993, with a brief stint doing comms for Microsoft. So she’s well aware that the first priority is to make sure people understand the ‘why’. The network isn’t pursuing AI for AI’s sake, but to enable its people to drive brand differentiation for clients by being the best ‘creative solutionists’ they can be.

“To drive AI adoption in the context of the future, believe in what our agency's going to do. Outline the vision, don't just talk about AI. I think that's really important. So people have the context of seeing a strategy and a vision and then see where AI fits in that,” she says.

For Dickon, AI represents a cultural, rather than technological change and will enable people to evolve into flexible, multi-specialists (‘M-shaped people’ rather than ‘T-Shaped people’). Both Patou and Dickon talk about ‘augmenting’ people. For it to be a true business-wide cultural change, it means inviting people from across the full range of departments to have a go and get hands on the tools and platforms.

“It’s a gain across everybody, that’s why we purposefully try to make it truly collective,” says Patou. In the early days of their AI journey, they got feedback that many people didn’t think that AI was for them; it was for the creatives who were instinctively drawn to playing with new toys and could make eye-catching demonstrations. Some teams even felt a bit left out. So in their training materials and communications campaigns, the leadership has deliberately made sure that they’re including examples and tips that apply to departments like finance or talent, to help them get excited about the ways AI could help them.

“It’s a mindset that you drive through embracing that this is an undeniable future, It’s an exciting future, it’s a culture. And we do it together,” says Patou. “You don’t want to leave anybody out.”

Individualised Journeys

Rather than try and force a one-size-fits-all approach, the leadership team started from an understanding that people will approach AI at their own pace, with their own level of enthusiasm and curiosity and bring with them their own concerns and baggage.

“[It’s important] to recognise that people are on different learning journeys, and that's OK, and to be actually quite vocal about that,” says Patou.

Roughly speaking, Patou sees three general categories: the self-taught super-user who is an early adopter and keen experimenter; the curious middle ground who need a bit of support getting to grips with the tech but who benefit from clear use cases and hacks; and then there’s last group, for whom the prospect of AI feels overwhelming and even worrying.

In order to meet people where they are at, the leadership deliberately created a multi-stranded approach to training, from giant webinars and creative inspiration sessions to pop up bars where people can go to ask specific questions. They’ve also been keen to mix the content up, balancing cheeky quick-win AI hacks with more substantive, in-depth use cases. As we’ll see later, they also made a point of encouraging a lively culture of peer-to-peer exchange.

Empathising with Reticence and Creating a Sense of Safety

Ogilvy has successfully taken adoption to 80% – that means 80% have used an AI tool. But, talking about individual AI journeys, there’s that last elusive 20%. Some of that is simply people who can’t be captured by the data due to things like turnover or being on leave, and there are others working on clients who mandate no AI to be used on their account. But, acknowledges Dickon, a fraction of that is people who just don’t like the idea of using AI.

In that case, empathising with the root of people’s concerns has been crucial. With strategy, initial resistance came when strategists feared that AI would be used to take away the very thing that they gained the most enjoyment and satisfaction from: digging into the research to unearth gems of insight. Instead, explains Dickon, the team was able to show how the tech would augment them, to do more of what they love. “You already get two weeks scoped in to define a project, but I'm going to give you the ability to, to consume a year's worth of intelligence,” he says.

There's been a deliberate effort to foster a sense of safety, helped by the guardrails of the WPP Open system, but also, a culture of experimentation.

There are others who worry about the ethical implications and long term impact of AI, and that by using it even in a professional capacity, they’re giving away too much of themselves. WPP Open has been created to be a safe platform, where security and data privacy are key, not least to protect clients’ sensitive assets and information. Despite fostering a sense of safety, to enable experimentation, and despite the extensive communication on the topic, there remains a group who don’t see the difference between a walled garden AI platform and the likes of ChatGPT and Midjourney, haven’t engaged with the training. At this point, Dickon says they haven’t wanted to mandate AI usage because it moves things to ‘a different level of engagement’ and it runs counter to the bottom-up wellspring of ideas and innovation they’ve seen.

And even those who have plunged into AI feet first aren’t without their own concerns or questions. To an extent, it would be worrying if people didn’t have uncertainties in a space that is so full of unknowns. For example, both Patou and Dickon reflect thoughtfully on some of the potential challenges of working with AI, such as the recent MIT research on the impact of AI on critical thinking. It symbolises an approach that is nuanced and granular, and that steers clear of one-note evangelism and hype. “Even the champions, people like me, we often feel overwhelmed by all of it. It is going so fast,” says Patou. That’s why they’ve wanted to create a conversation, internally, and also what inspired them to create their internal advertising campaigns.

Helping Clients on Their Journey

It’s not just the handful of team members who are reticent about AI. Patou says that Ogilvy’s clients cover the full spectrum of AI attitudes, from true believers to those who are 100% against using it. Part of the education on AI includes legal and ethical training to make sure that brands are safe. It’s also crucial that there are guardrails in place and a culture of transparency to ensure that every client has just the level of AI touching their brand as they are comfortable with and have mandated.

Having overseen the AI transformation within Ogilvy, Patou says there’s a huge opportunity to transfer that knowledge to clients and help those who are looking for help.
“I think clients are on their own transformation journey,” says Patou. Aside from using WPP’s suite of AI tools to help clients with their campaigns, Patou sees an opportunity for Ogilvy to combine its consultancy and behavioural science expertise with its hands-on experience activating its own AI transformation. Whether clients are looking to educate their employees or drive their own culture of AI adoption, she sees a real role for the network to share its insights and help with the finer points of change management.

Experimentation to Adoption to Cultural Saturation

According to Dickon, instead of mandating AI usage, the team encouraged people to experiment with the tools personally.

“The critical thing is, if you need to change the culture around adoption, you need to let people experiment,” he says. And that’s not just on business problems. Allowing people to apply the tools and training available to their own personal needs (creating a recipe or helping devise an exam study schedule with their kids, for example) also creates a safe space for them to learn without worrying about any business impact. Then it becomes easier for people to figure out how to transpose what they’ve learned to a professional setting.

I think that's that's where we are now,” reflects Dickon. “That journey has got to a point where people we're trying to shift it from a technology phenomenon to a cultural phenomenon.”

Local Heroes and Peer-to-Peer Learning

Pushing a cultural and technological transformation in a top-down manner was never going to work, particularly considering the diversity within Ogilvy’s EMEA region, stretching as it does from Stockholm to the tip of South Africa.

Each market has its own needs, skillsets and contexts to account for. In the big markets like the UK, France and Germany, the scale brings lots of opportunity and plenty of challenges, whereas in “historically more innovative” markets like Poland and Central Europe there’s been a real sense of hustle and smaller teams that allow for faster experimentation. In Albania, Dickon says there’s a “supremely technology first” attitude, which is borne from the country’s relatively recent government and private sector-driven tech advancement.

Meanwhile, in the Middle East, clients are coming up against the fact that most large language models are built on English. Organisations are building their own Arabic LLMs and local offices are figuring out how to innovate within that ecosystem, as opposed to ChatGPT or Google.

Reflecting on markets like the UAE and Qatar, Dickon says, “They are very technologically-hungry and growth-hungry markets and they are accelerating because all the businesses are wanting to do more. Naturally, our offices are primed to stick with the clients and the culture and the governments that are leading it out there.”

That’s why, for Patou, encouraging local-level leadership and grassroots exchange and knowledge sharing has been key to fostering relevance and therefore cultural change.
“I think the most important thing in learning and development is that peer to peer exchange, basically,” she says. “It is through, I think, a sense of “people like you” who can demonstrate how they've discovered it, what the use cases actually are, how AI can help them in a task or project that you actually recognise. That is where we see the fastest traction.”

Tracking Impact and Building New Workflows

With a clear idea of what AI is in service of – brand differentiation and augmenting people – and a wealth of hacks and use cases to draw from, Dickon has been keeping an eye on where the technology is effective and where it isn’t, and which tasks can be automated and which should be augmented.

As Dickon and the team figure out where to fit AI into standardised workflows - say, the steps of onboarding a client – they need to analyse where it will be most effective.

“We have a process whereby whenever we're designing one of those workflows within one of our team, we map the existing business process that's been used for a decade or even a century in terms of getting work through an agency. And then we work out where are human beings actually more valuable than AI? [Tasks involving] empathy and high judgement, and what tools do they need to do that job? We call it the empathy gradient,” says Dickon.

“On the other side of that, where do humans really cause a bias?” Financial forecasting is one such area, where pressure from CFOs or ideological bias can really skew humans’ interpretation of data.

It will be a relief for many to find that AI still can’t come up with a decent creative idea. “It was very, very literal,” recalls Dickon. “It won't make the jump between the reality and then the mental model that a human being has of what life might be like. They can't embed empathy into it. So that's the drop off… As soon as you start automating the creative process, the performance goes through the floor. And we've obviously experimented with it because it would be weird not to.”

Many of the early wins of the roll out involved taking on the admin tasks that bog down project and account management. With the growth of agentic AI, people have started to realise that agents can take on the work formerly done by executive assistants… a role that many agencies haven’t had around for years.

The creation of digital twins to mirror audience cohorts has been another game changer - ideas can be tested instantly, with an accuracy of around 70%. Less than a human focus group, but much faster.

As use cases and hacks flood in from across the region, Dickon’s new challenge is to keep track of them all. Whether it’s turning a dry company annual report into a lively podcast to listen to en route to a meeting or finding ways to automate the evergreen pain point of timesheets, the ideas are coming thick and fast.

What’s Next?

As much as the Ogilvy EMEA team has made considerable progress, there’s no time to kick back and relax. For Dickon, the big journey now is moving people from adoption, which is a fairly binary measure, to more mature, rounded usage. Maturity, for Dickon means understanding the true impact of AI and continuing to facilitate a culture that embraces it.

Whatever the ongoing AI adventure throws up next, Patou is confident that, thanks to their innate creativity, Ogilvy will be able to face the challenges and embrace the opportunities.

“That is creative agencies’ superpower, because we’ve had to pivot so many times and have to stay at the front,” she says. “So, no matter what comes our way, our ability to deal with it, to lead it, to be curious around it, to understand it, basically, and to help our clients with that is probably the uber muscle beyond all this.”