I’m having more conversations than ever with marketing leaders about AI. The interest is high, the pressure is real, and the feeling that something needs to change is unmistakable. What’s less clear for many teams is where to start.
That tension was at the heart of our recent panel discussion on building the foundations for AI enabled marketing. Despite the focus on AI, this wasn’t really a conversation about tools. It was about the organisational shifts required before AI can deliver meaningful value.
The uncomfortable reality beneath the AI hype
One of the clearest themes was that AI is exposing problems that already exist. As Emma Moorman (Informa) shared, the tool itself is only a small part of the equation. If data foundations are weak or workflows are unclear, AI will make the gaps more visible, more quickly.
Sharon Roessen (Terrapinn) reinforced this with a practical example. Their initial move to AI enabled sales tools didn’t immediately improve performance. Instead, it highlighted inconsistencies in process and adherence to methodology. The real work became redesigning workflows, documenting ways of working and retraining teams.
What I see repeatedly is teams starting with tools because it feels like progress. The progress that sticks comes when organisations step back and understand how marketing actually operates today, and what needs to change.
Governance as a source of confidence
Governance featured heavily in the discussion. What stood out was a shared reframing of governance as something that enables movement rather than restricting it.
Sharon spoke openly about the risks of uncontrolled AI usage, particularly around data. Limiting access, defining ownership and putting practical guardrails in place became essential once AI made it easier to move fast and at scale. Without those guardrails, mistakes simply happen quicker and cost more.
Emma described a similar approach at Informa, where internal AI environments and clearly defined use cases create safe spaces for experimentation. The aim is not to slow teams down, but to give them permission to move with confidence. Without that clarity, people find workarounds and governance disappears entirely.
Capability is built through real work
Another strong message was that capability isn’t built through abstract training, but through application.
Emma shared how the most effective learning happened when teams worked on real problems together. Live campaigns, real planning challenges and genuine customer journeys created far more confidence than generic AI training ever could.
Sharon added that removing risk is critical. At Terrapinn, centrally run pilots allowed value to be proven without putting pressure on individual teams or budgets. Once impact was demonstrated, scaling became much easier. Capability grew because people could see what worked and why.
Redesigning workflows unlocks value
Some of the most compelling use case examples focused on planning, data and decision making.
Emma described how AI supported scenario modelling and pricing decisions, enabling more confident strategic planning. She also shared an example of personalised brochures, where AI allowed a level of personalisation that simply could not be scaled before. The value came from improving the customer experience, not just saving time.
Sharon’s examples focused heavily on data. By redesigning how ideal customer profiles were built using AI enriched historical data, Terrapinn automated a previously manual and debate heavy process. The outcome was better targeting and stronger commercial focus, not just efficiency gains.
Experimentation with intent
Experimentation still matters, but the discussion drew a clear line between useful experimentation and endless testing.
The teams seeing progress define success upfront. They are clear on whether an experiment is intended to drive efficiency, revenue or learning. That clarity makes it easier to scale what works and stop what doesn’t.
There is still value in smaller, local experiments while organisations build maturity. The key is being honest about purpose and expected outcomes, rather than treating experimentation as an end in itself.
Follow up insights from the Q&A
We didn’t get through all the questions live, so I wanted to share a few follow up insights that expand on the discussion.
On building and driving adoption of an internal LLM
Informa’s internal tool is built on a range of best in class third party LLMs and agents, including Claude, ChatGPT and Deepseek. Users can toggle between models, which are updated as new versions are released. Adoption was driven through a sustained programme of sharing, roadshows and learning together across the business. The focus was not just access, but confidence in how to use the tools well.
On change management and shifting fear into action
The shift from fear to action happens when change is made tangible. Not more awareness, but a clear first step people can take in their own role. This principle applies just as much to AI adoption as it does to climate or sustainability initiatives. Meet people where they are, show early wins and build momentum from there.
On defining success for pilots and proofs of value
Success needs to be defined upfront. Emma frames this in two outcome buckets: efficiency and revenue. Clear KPIs within those outcomes, and agreement across stakeholders on what success looks like, are essential.
On favourite AI use cases
Emma is most excited by use cases that unlock personalisation at scale. Personalised brochures were highlighted as an example of something that simply was not possible before, but is now delivering meaningful customer value. Sharon focuses on how Terrapinn automated a previously heavy manual process by redesigning how ideal customer profiles were built using AI enriched historical data.
The foundation that matters most
The biggest takeaway for me is that AI enablement is not a tooling challenge. It’s an organisational one.
Confidence comes from practical enablement. Capability is built through real application. Clarity comes from strong foundations across data, workflows, governance and leadership. Tools will continue to change quickly. Those foundations are what allow teams to adapt without constantly starting again.
The opportunity is here. The question is whether your organisation is building the foundations needed to take advantage of it.
By Zoe Merchant, Partner – Growth, Marketing and Sales Consulting