Personalised communications have long been a dream just out of reach for marketers. With AI, that promise is closer than ever, though with one proviso: your data needs to be ready.
“The opportunity is wide open for everyone in this room,” Bronwen Edwards, director and head of martech at Credera, told industry leaders at Tealium’s recent Building the Org of the Future event in Melbourne. “Personalisation was hard before, but AI has raised expectations and accelerated the pace. We’re not just building customer journeys anymore, we’re building living, learning, breathing systems. Real-time personalisation isn’t just hype.”
The Data Imperative
AI’s promise is only as good as the data behind it. “Bringing personalisation from vision to reality is not simple,” Edwards added. “Only a small percentage of CX professionals are truly confident in their efforts. You need excellent data, strong governance, and a solid strategy. It all sounds simple in a PowerPoint, but executing it is complex. Despite all the LinkedIn hype, adoption of AI is slower than many expect.”
Challenges include legacy systems, poor data quality, limited experimentation time, and the difficulty of building a business case that wins company-wide support.
This reality was echoed by Yuuka Hama, digital CDP and solution engineering manager at Bupa, who has been responsible for fixing data gaps, improving identity stitching and standardising tracking ; the foundational work required before AI can deliver anything meaningful. Her takeaway captured the sentiment of the room: “AI runs on data. We cannot run AI when we don’t have data.”
“The number one thing you need is a data foundation,” said Ross Macrae, Tealium’s principal AI strategist. “That includes unified customer graphs, data from multiple sources—including second and third-party—and even environmental data. AI-ready infrastructure makes that data actionable in real time, giving agents the information they need to deliver personalised experiences.”
For Bupa, as AI and personalisation ambitions expanded, its existing CoE became essential to scaling responsibly. The model ensured consistent data practices, shared learning and coordinated experimentation, all crucial for reliable AI. “The CoE structure allows us to share learnings, maintain data integrity and drive collaboration,” Hama said. “Smaller businesses can move faster, but a central framework helps us scale efficiently.”
Tealium ingests data from web, mobile, CRM and enterprise systems (e.g. call-centres or data warehouses), and—via integrations with platforms such as Braze, Snowflake, Salesforce, Adobe and AWS—enables real-time activation and personalisation.
The AI Advantage
To achieve personalisation at scale, organisations need to understand and deploy large language models (LLMs). “These are transformational,” Edwards explained. “They allow us to move beyond basic segmentation and create interactions that are natural, adaptive, and dynamic, more conversational and human.”
Rather than relying on a single LLM, different models serve different purposes: Anthropic excels at reasoning, Gemini at context, ChatGPT at synthesising data, and Perplexity at deep research. With Tealium, Credera can organise vast unstructured customer data and turn it into actionable personalised communications.
Chatbots are evolving, too. “With long-term memory, sentiment, emotion, and intent analysis, chatbots can function like AI agents,” Macrae said. “With structured, accessible data, this can be a powerful first step toward AI-driven customer engagement.”
Data Orchestration in the World of AI
Real-time personalisation powered by AI offers clear advantages, but it relies on more than advanced models. It depends on a unified data foundation and the ability to orchestrate that data consistently and in the moment. Without strong data strategy and orchestration, AI simply can’t deliver meaningful, real-time experiences.
As Edwards said, real-time personalisation is not simply hype.
“It’s about orchestrating trusted, timely and transformative data flows supported by Tealium. You don’t need a magic AI engine, you just need a really solid foundation. Getting that foundation right is critical and the decisions you make today can either set you up or slow you down in the next 12 months,” she continued.
Giving an example of where poor foundations can lead to shaky builds, Edwards pointed to the myriad of data sources inherent in paid media buying.
“You’ve got multiple platforms, audiences, campaigns. Some are overlapping. But if you haven’t mapped your customer journeys and connected all the dots–the clicks, interactions, conversions–and how they fit into the broader customer experience, you’re missing out on feeding the AI the full story. You’re feeding it a jigsaw puzzle with pieces from different boxes,” she said.
Amin Foda, director of marketing infrastructure at Monash University, added: “Think about why you’re doing things: you’re doing it for the business, but without your audience you have no business. Behaviours are shifting, and organisations need to adapt quickly.”
A clear data orchestration strategy puts customers at the literal centre of business operations, creating a flywheel for growth as AI proficiency becomes a competitive differentiator. And as organisations mature, it’s not just the technology that must evolve, it’s the people.
Chris Harley, GM of data and tech at Annalect, noted that building AI capability requires sustained investment in skills and change management. For every dollar spent on AI technology, organisations often invest three more in training to ensure teams can use it effectively.
As AI agents become more common, it becomes clear which businesses have orchestrated their data correctly.
“The number one thing you need is a data foundation. In that data, you need unified customer graphs. We need to be able to link data together. We need data from multiple sources, even second and third party data, even data that doesn’t relate to the individual, it’s environmental,” said Macrae.
“I need to have that data to hand and we need to have the AI-ready infrastructure which means it can be real-time and agents can access the data and bring it to the fore.”
As the market becomes more sophisticated, bringing that data to the surface will not only be a competitive advantage for your brand or business. It may also become the very foundation of your organisation’s survival.
