DIGITAL TRANSFORMATION SERIES - Using AI to supercharge our teams

My purpose is to unleash human potential to create a healthier world. My vision is to inspire pharmaceutical commercial and marketing leaders to transition from merely delivering messages to becoming trusted advisors to healthcare professionals. My goal is to unlock the power of genuine connection. I am passionate about helping pharma professionals feel confident in using technology to improve their engagement with healthcare professionals. When it comes to digital transformation, I am particularly keen on enabling individuals to use technology with confidence, an area that fascinates me. Over the past two years, I have explored this topic by interviewing inspiring pharma leaders to learn from them and provide a platform for them to share their perspectives.

I was delighted to interview Marco Andre, Global Head of AI and Marketing Excellence at Novartis. I was particularly excited because Marco is quite unique. He possesses extensive experience in AI and automation, yet he brings a refreshing humility to how these tools can be applied in the pharmaceutical industry.

Marco is a keynote speaker, an AI advisor to executives, and a published author. With over 20 years of experience in companies such as Google, Procter & Gamble, and YouTube, he has held local, regional, and global roles in marketing, sales, partnerships, and operations. Outside of work, Marco is the author of the book Imperfect Stories, a passionate Lego enthusiast, and a proud owner of more than 20 sets.

I believe I first met Marco in person in Dubrovnik in May this year, where he delivered two memorable presentations. On Day 1, he gave a talk titled "The Hype is Gone, It's Time to Grieve," focusing on generative AI. His presentation was so entertaining that I found myself sitting at the edge of my seat; it was fast-paced yet highly engaging. I think he received the highest score for the upfront presentation that day. On the second day, he led a brilliant interactive panel discussion about learning from failures. I thoroughly enjoyed his sessions—they were fun, interactive, and refreshingly real.

Let’s start with your career, which spans a range of industries from consumer goods to digital marketing, and now to pharmaceuticals. What key lessons have you learned from your time at companies like Google and Procter & Gamble, and how have these shaped your approach to digital transformation in the pharma industry?

It’s interesting because, in a corporate environment, you have a choice: do you want to be a rebel? If you’re a rebel, you’re probably better off on the outside. Alternatively, you can choose to bring people along with you. There’s a myth that everyone loves change, but that’s simply not true. Change can be quite challenging. A privileged minority may embrace it, but the reality is that most people find it painful, even if they don’t mind it.

One important lesson I learned in these companies was that I didn’t want to be a rebel; I wanted to be part of the system, managing it while driving change and innovation. Digital transformation is just one aspect of this, and now we’re experiencing a wave of AI. The second thing I’ve always tried to do is to include everyone—not just for the sake of inclusivity, but because it genuinely yields better outcomes. Engage with the most senior and the most junior people, not just within your own function, but across HR, finance, and sales. The more people you involve, the more they can see the shared value.

The third lesson is recognising that we live in a time and a world that is constantly evolving, which I illustrated with a metaphor in Dubrovnik. We often try to convince the waterfall to behave like a lake, but the waterfall will not stop. In companies, it’s crucial to understand that waiting for a moment of stability is futile; change is the default state. At a certain point, I realised that adapting to change is essential, and I even learned to champion it.

What lessons did you learn from working in previous industries, and how can we incorporate them in pharma to ensure everyone is included in the change journey?

Size can be both a blessing and a curse. When I joined Google, it had around 20,000 employees; ten years later, it grew to 300,000. Initially, it was relatively easy to maintain close connections in the global marketing department, which had 300 people when we started. As the company expanded, however, it became incredibly challenging, leading to a greater reliance on culture rather than just people.

Many companies try to tackle this with processes, which can be useful, but culture is far more impactful. Processes can never be exhaustive; if you have a strong culture and guiding principles, people will instinctively know how to act in various situations. I recall a former CEO of a UK retail brand who said that the only thing a store manager couldn’t do was throw a party; everything else was up to them. Creating that culture and those guiding principles takes effort, especially as a company grows.

If you rely solely on processes, they will eventually fail. It’s essential to give people the autonomy to make decisions. At Procter & Gamble, we had something called Purpose, Values, and Principles (PVP). During difficult discussions, people would refer to these principles to evaluate whether a decision aligned with our PVPs. It’s vital to implement these principles in practice, rather than merely displaying them on a wall, which unfortunately happens in some organisations.

Do you think we have that culture in pharma? Do you think there is a culture that allows people to make decisions and follow guiding principles?

The fact that pharma is a regulated industry does make it harder to create such a culture. However, I’ve seen companies excel when they recognise that, within a certain regulatory framework, they can encourage people to rethink processes through the lens of technology. A common mistake, not just in pharma but in other industries too, is to apply technology to broken processes. The result? Those processes appear even more flawed.

For example, if we’re producing content in the pharma industry and manage to do so in a third of the time and cost, it’s pointless if we then submit it for MLR (Medical Legal Review) that takes the same amount of time as before. We need to think in an integrated manner. What I always emphasise about AI in pharma is that while there is a regulatory framework and existing processes, we must rethink how technology can disrupt and improve those processes, enabling us to create better and more efficient workflows.

In your role at Novartis, can you share a bit about your responsibilities without revealing any confidential information? I’m curious to learn how you lead in a way that others can benefit from.

For the past two years, I started by overseeing all marketing capability training across the company. Shortly after I began, around December two years ago, I noticed the rise of AI and felt compelled to explore its implications across the company. I started by reaching out to various functions and countries, aiming to educate them about AI — how it’s transforming not just marketing, but every function.

My role has involved training people about the implications of AI on our go-to-market strategies and our interactions with patients and customers. Over these two years, I’ve worked to raise awareness and provide capabilities at scale, including training senior teams. It’s been fascinating, as two years ago, when I first approached teams with this concept, many thought I was crazy. Today, AI is much more on people’s minds, and there’s a growing understanding of its possibilities, with fears diminishing.

How do you see AI reshaping not just marketing, but the entire business model for pharma companies like Novartis?

When discussing go-to-market strategies, we must start by observing where the market is heading. With the emergence of these tools, both patients and healthcare professionals (HCPs) are evolving into super patients and super HCPs. They will have unprecedented access to information at their fingertips, fundamentally changing how they operate.

Reflecting on when we began transitioning to mobile technology in the early 2000s, many dismissed the idea that the world would go mobile-first. Yet today, people spend far more time on mobile than on desktops. Similarly, we are witnessing a shift where, for instance, a doctor might quickly access information through a tool like GPT instead of navigating an HCP portal—remembering URLs, logging in, and searching for data.

The same applies to patients, who will become significantly more informed. I once heard the story of a woman who consulted 17 doctors over three years before diagnosing her son’s condition by inputting his exams and notes into ChatGPT. This highlights the power of informed patients who will pose challenging and insightful questions to HCPs, compelling the pharma industry to be agile and provide information in a way that resonates with them.

On the commercial side, we can expect substantial changes. In research and development, advancements like AlphaFold are likely to accelerate development cycles by enabling extensive in silico testing. However, I believe the most profound impact will be felt on the commercial side, where we are already seeing significant shifts.

How can we use AI to bridge that gap so that it can provide relevant information quickly, ensuring that the information is available when people are searching for it? How do you think we can overcome that?

If we take the example of this conversation we're having right now, in another context, we would finish this call, take notes, and turn this into some form of written content. Right now, what is happening is that we are getting a transcription from this call at the touch of a button. This will provide us with insights that are already valuable because it won't be biased information. It won't just be what you and I remember; it will be pure transcription. From that, we can create a piece of content, probably using a GPT model, which, of course, a human will need to review because we want to avoid hallucinations or misquoting what you or I said. Taking this into pharma, it’s exactly the same. You can use these tools to accelerate parts of the process, knowing that a human always needs to be involved. From this small example, we can already see that what we perceive as bias, something we know AI can propagate, may actually be less detailed and incomplete than the notes that you or I would write.

How can pharma inform patients directly with evidence-based content to help them become empowered patients?

I think several factors are at play that we all need to be aware of. A lot of what patients are looking for—or even people like us—has been searched for on Google. The concept of Google Search is to find what is called the 'right answer'. However, with large language models such as ChatGPT or others that patients might be using, there is the concept of ‘no one right answer’. The goal is not necessarily to find the absolute right answer, but the answer that is right for you, based on what is known about you and the context you've provided, along with the most up-to-date information.

Therefore, what will be interesting for us as an industry—and this applies beyond our sector—is if something is being searched for on Google. Currently, we have some form of control through SEO. Yes, we can do search engine marketing, but in Europe, we can use SEO to try to manage what appears. There will soon be a new discipline that could be termed something like LLM optimisation. If a patient or a doctor is looking for information on a specific drug, we need to ensure that the information is accurate; otherwise, there could be significant repercussions.

Traditionally, in Europe, we are very good with regulation. Europe was one of the first groups of countries to develop the AI Act. My concern is that while we implement a lot of regulation, it doesn’t leave much room for innovation. So the answer is, I don’t know, but I see shifts already happening.

Marco, for the benefit of everyone, could you explain what LLM stands for?

LLM stands for large language model. Generative AI is a subset of AI that often employs deep learning to create new data based on existing patterns. A large language model is a type of generative AI focused on understanding and generating human-like text. An example of a large language model is ChatGPT, where users pose questions and interact with the model to generate responses.

You can't really stop patients from seeking information from other sources, can you?

No, I don’t think you can. The sooner we acknowledge this, the sooner society will understand where things are heading and can prepare for it. I often reference the millennium bug on 1 January 2000, when we knew that at midnight on 31 December 1999, everything would falter because computers were not prepared for the date change. We had time to prepare; we established procedures and updates, so we managed it. My concern with AI is that if we remain in denial, believing this will just go away, we are not preparing and are missing opportunities—not only as businesses but as a society—to adapt to where the market and patients are heading.

It’s like trying to regulate something you can’t even control, isn’t it?

You've worked with billion-dollar brands and led digital marketing before it was mainstream. How would you compare the challenges of digital transformation in pharma with other industries like consumer goods or technology?

I think the nature of pharma is slower, and we all understand this for very valid reasons, largely due to regulation. In technology, a minor update to a YouTube algorithm doesn’t impact lives. In consumer goods or food, the stakes are higher, but in pharma, it’s even more critical. That's the main difference. People often say that pharma has much to learn from tech, and I agree, but I also think tech has a lot to learn from pharma. Technology is becoming increasingly regulated, as it should be, and could benefit from pharma's experience in operating within a regulated environment.

My hope for pharma, and the reason I transitioned from tech to pharma, is that even a 1% increase in speed to market can have a huge impact. It can mean getting a drug to market faster, which may save a life, improve quality of life, or extend life. Thus, I advocate for a mindset of moving slightly faster and testing new approaches to discover new outcomes. I believe AI is an excellent means to achieve this.

A comment from the audience. Stefan states there is an urgent need to ensure that relevant information is considered by the models in real time. What would you say to that comment?

He’s absolutely right. A lot of what we currently don’t know and feel uncertain about stems from how these models are trained on various data sources. While we are beginning to understand the data they are based on, the processing and how it is ultimately presented to patients, consumers, and professionals remains somewhat opaque. At some point, this needs to be clearer, as it relates back to search engine optimisation. For many years, we understood the basics of how the model worked, but not entirely. There were large companies trying to navigate it without access to the algorithm, yet they knew that certain practices would lead to good SEO positioning. I think a similar scenario will unfold with AI, and Stefan is correct—we need access to that information and clarity on how it is processed and presented.

You mentioned that pharma is heavily regulated and has a traditional approach. How do you navigate the balance between innovation and compliance when driving digital transformation initiatives?

What I always advise is to take the example of AI. Sometimes in pharma, we assume that if we're discussing X, we mean we want to implement X and send it to market tomorrow. That’s not the case at all. Many of us are already thinking about customer- or patient-facing AI tools, which are obviously subject to regulation, and adapting to that will take time.

I always encourage leaders to think internally first. Consider how you can use AI to enhance productivity, time, and the capabilities of your teams, enabling them to perform better or enjoy a better quality of life. For instance, many corporations suffer from 'corporate amnesia'—when someone leaves, vital knowledge is lost, and documents are scattered across numerous locations. Much of this can be alleviated with AI. This doesn’t mean we will apply the same approach to our customers or patients, but starting with our internal teams is essential. If you have an objection handler, you could use some form of GPT to centralise that knowledge, allowing the same people access to the same information. Of course, there must be some human oversight to mitigate hallucinations. The goal is to supercharge your teams. That’s how I believe we should implement AI in pharma: start internally, and the external use cases will follow.

Another question from the audience. Laura asks, Do you think AI can assist us with strategic choices so we can focus our activities better?

Absolutely. AI can significantly aid in strategic work. There’s a component of AI that can handle tasks such as writing emails, performing Excel functions, or creating presentations—all of which I refer to as 'brute force work'. We often spend 20-30% of our time on these tasks, and freeing that time is valuable.

I also see substantial value in making strategic choices. For example, I came up with a narrative regarding the biases and pitfalls of AI, initially during a plane journey. I wanted to call it the "cardinal sins of AI" and had some ideas, particularly around hallucination and bias. During a two-hour drive, I conversed with ChatGPT, discussing my narrative and asking for pros and cons of hallucination. It helped structure my thinking and suggest other relevant topics.

AI tools can serve as a sounding board for discussing strategic decisions, offering a perspective that can help organise your thoughts in ways you might not achieve quickly without assistance. I relate to this personally; I have a business coach, and sometimes when I face dilemmas or strategic questions, I bounce ideas off them. Since I began using ChatGPT, I realised I could resolve some quick questions without needing to consult someone else.

However, my question revolves around data privacy. How secure is our data when we interact with AI?

It largely depends on the model. For instance, certain large language models, such as ChatGPT or Claude, have features that allow you to opt out of having your input used for training purposes. What we know now is that all data available online—from Reddit to Google—has been used to train these models. Recently, LinkedIn introduced a default opt-in setting for everything users write on their platform to be used for training, but they couldn’t implement that in Europe due to regulations.

The key point is: if we sacrifice our privacy, what do we gain in return? This was the same concern with the initial wave of digital technology. If we realise these models are trained on our data without any benefits for us, adoption may slow. I believe there must be a balance: we can provide our data to train models but should receive something back—such as strategic insights or reduced manual tasks.

How do you think pharmaceutical companies can use storytelling to better connect with healthcare professionals and patients in the digital age? Does AI play a role in that?

It's interesting because I saw a clip the other day—although I can't remember the name—but they were asking someone what we should teach our children for the future, as knowledge is becoming more accessible. The answer came from Scott Galloway and was quite enlightening: storytelling and communication skills are crucial. The ability to articulate a vision and create a shared understanding among people will become essential, especially as healthcare professionals (HCPs) and patients gain unprecedented access to information. What will differentiate us in this new world is empathy, communication, and the way we tell stories. I mentioned a provocative statement from a billionaire earlier: he suggested that in six to ten years, we might not need doctors, but rather actors, because knowledge will be universal. While I don’t believe that will happen, the emphasis on storytelling will become even more important than it has been until now.

What are your thoughts on generating trust and creating human emotions in this context?

I believe it’s all about how we perceive AI. For me, AI represents a set of superpowers that enable me to perform my job better. For instance, with tools like Salesforce, I can enhance my preparation for interactions. If I start receiving deeper and more insightful questions from HCPs, I want to be well-prepared. AI should work in the background, helping me improve my performance rather than replacing me. This applies to sales as well.

How do you help senior pharma leaders adopt and embrace digital transformation, especially those who are hesitant about new technology and AI?

Many issues are internal at the senior level. I advocate for getting them into a room together to engage with the technology directly. They need to experience it firsthand—interact with laptops and use the tools. Ethan Mollick, a well-known professor at Wharton, has pointed out that the biggest barrier to the adoption of large language models is senior executives not using the technology. What we fear often stems from what we don’t understand. This is not about executives being resistant; rather, in a world where they are pressured to deliver results, AI can seem like a luxury. By demonstrating how quickly they can create a video, for instance, we can shift their perspective and encourage them to consider their roles, potential partnerships, and how to integrate this technology into their operations. Society and businesses often fail to invest adequately in developing capabilities and ensuring people understand technology. We focus too much on the tech and data at a corporate level without truly understanding it. That needs to change.

Let’s talk about your book, "Imperfect Stories." You focus on the beauty of imperfection, which I find particularly relevant in pharma, where perfection is often sought. How can embracing imperfection and a culture of experimentation benefit pharma companies navigating digital transformation?

The impetus for my book stemmed from how business and education condition us to believe we must have everything figured out. The rise of social media over the last 15 years has amplified this pressure. However, we see more stories now of people who haven’t had it all figured out. In the pharma industry, this pressure carries more weight since lives are at stake. While we must respect that, we can still acknowledge that technology has pros and cons. For example, AI can propagate bias, but it’s humans who create that bias through their decisions. The data used to train AI reflects those biases. However, we can also use technology to address and dismantle biases. The story of Viagra illustrates this; it began as an adverse event but became a life-changing drug for millions. In pharma, we need to recognise that there are cons, but also weigh the pros and explore how to use technology to our advantage.

Another question from the audience. Eva asks how HCP engagement will change, particularly regarding communicating knowledge about new products and treatment options. How can this shift to storytelling and empathy while keeping it interesting for HCPs?

I believe this will be incredible but also quite challenging. HCPs will have access to the information they need in their preferred format and context within seconds. Future HCP engagement will hinge on our ability to quickly respond to their inquiries. With large language models, we can generate insights and relevant information much faster than before—potentially in days or even hours instead of months. However, from an organisational standpoint, we are currently ill-equipped to achieve this. If I were discussing this with Eva now, I should be able to generate a document addressing her question with various scenarios regarding HCP engagement in no time. This is where we need to focus our efforts.

What do you love about Lego? Does it help you unwind? Tell me more about your passion for it.

It truly helps me unwind. There’s been research suggesting that many people in various jobs feel unhappy not because they dislike their work, but because they don’t see the tangible results of their efforts. Lego allows me to build something with my hands by following instructions, requiring less cognitive effort from me. I may not be particularly creative with Lego, but I find fulfilment in seeing the finished product. Each piece feels like it belongs to me, which is why I spend a considerable amount of time with it—it provides a connection and allows me to unwind away from screens and the demands of daily life.

You read "The Omni Advantage" and shared your feedback with me. What are your thoughts on it? I believe we need to empower our people more, focusing on the human element of digital transformation, rather than the technology itself. What are your views on change management in the pharma industry to engage more meaningfully with HCPs?

One interesting aspect for me was the chapter discussing the shift from message bearers to trusted advisors. You wrote this just as we were entering the AI era, perhaps not realising how much AI would enhance this transition. HCPs will have instant access to knowledge, so the messages we deliver need to be meaningful. Ultimately, trust becomes paramount, achieved through empathy, storytelling, and usefulness. While we may have great stories and empathy, the depth of content is essential. Your insights were quite forward-thinking, and I believe we must prioritise content and the quality of relationships over merely disseminating messages that HCPs can access elsewhere.

Conclusion

Marco's insights provided a wealth of knowledge that was both enlightening and thought-provoking for the audience. His humble approach and genuine interest in people truly shone through, emphasising the importance of storytelling, empathy, and a culture of experimentation in the pharmaceutical industry. The discussion highlighted how these elements could significantly shape the future of pharma 

If you would like to purchase a copy of The Omni Advantage, it is available as a paperback or audiobook.