Interview with Graham Addison, Senior Director for Global Launch and Marketing Excellence, at AstraZeneca’s Global Hub in Barcelona.
In this exclusive interview, SciencePOD founder Sabine Louët talks with Graham Addison about AstraZeneca’s Digital Twin initiative, which promises to yield deeper insights into how best to anticipate and respond to HCPs’ individual needs.
Addison will deliver the keynote ‘Digital Twin: The Human Touch Supercharged’ at Ubiq Events 4th Pharma Omnichannel HCP Engagement Conference in Barcelona, running 10-11 March 2026.
Could you tell us about your current role?
I’m responsible for Global Launch and Marketing Excellence at AstraZeneca. For example, we define the strategic focus and own the process of strategic planning across 22 brands. And with Ambition 2030, we are on track to launch 20 new medicines, become an €80 billion company, and be carbon negative.
One of the most exciting projects is the launch of AstraZeneca’s Digital Twin initiative, which is currently running across Europe and Canada and which I’ll be talking about in Barcelona at Ubiq Events.
Could you describe what the Digital Twin initiative entails?
Digital Twin is an AI model that allows us to create a digital replica of a healthcare professional, so that we can put the customer at the centre and truly address their needs.
Digital Twin works by creating a ‘digital replica ’ of each healthcare professional. It uses an AI model to predict their behaviour. The model allows us to understand to n=1 recommendations for optimal channel mix and frequency for each healthcare professional. This process considers various data inputs like characteristics, interactions, sales and market share data, and market access information.
We only launched in September last year and are still refining the model, making sure that we have the correct datasets and, of course, are getting the correct outputs.
What is the nature of your work in this project?
My role within Digital Twin is fully centred on change management, so I’m part of the core project team. But my role specifically is around driving the organisational change needed to implement that strategy. The actual implementation phase of our Digital Twin into a new market takes 14 weeks, so we do it in seven sprints, and it’s this setup stage that’s taken up the majority of our time.
What kind of value does Digital Twin add to your understanding of an HCP?
Any data that allows us to better understand our HCPs or doctors, wherever they might be: It will take into account congresses and meetings they’ve attended, emails they’ve interacted with, anything they’ve opened, where they’ve spent time, how many times they’ve seen a representative, how many times they’ve interacted with specific channels, and whether it’s digital, in person, virtual. We also look at where they’re focusing and spending time in relation to key therapy areas or products.
We’ve started small with this product and are trying to scale up fast. We began with four markets back in September, and now we’re up to nine. We’re only focusing on certain brands for now. The real value for us will come when we expand to include more outputs and brands.
Digital Twin obviously has an impact on our ways of working and how we conduct our business. But from a marketing perspective, you tend to have the interests of your brand at heart. When you’re on a brand team, your product is the most important thing to you.
If you think about what I’ve just described, that really is putting the customer at the centre, going to see them with the product we know they want to hear about, which will interest them, and which meets a need we know they are experiencing.
How much insight can you get from the Digital Twin in terms of informing your marketing activities or even your MSLs activities?
If you think about historical segmentation and targeting exercises, you’re bucketing physicians into categories A or B. This removes that broad grouping and gives us an individual-level call plan.
One of the things that I’ve been working on with the different countries is taking the outputs from the model and integrating them into the call-planning cycle. The end user for this is the sales representative who then gets that call list or target list. So, there’s a management process involved in that.
The next layer up is the marketeers, who can then access omnichannel microjourneys content, targeted content, and personalised content. Using the Digital Twin model, for specific doctors, we can understand their medical content consumption preferences, and their behaviours around channels, formats, and timing when accessing our medical information.
The final piece in the puzzle is a better understanding of resource allocation and optimisation.