CEO & Co-Founder, Digital Health Coalition (DHC)
How does the pharmaceutical and medical device industry think about AI today? How is that changing over time?
The Digital Health Coalition recently conducted research with brand and marketing executives at approximately 25 global pharmaceutical companies and posed that very question. What we learned is that while AI is not at the top of their strategic radar today (in 2018), their likelihood to rank it as a very high priority in 2019 significantly increased. In other words, it’s not impacting their business today, but they are very much aware of the fact that AI is quickly permeating all sectors of our society and healthcare is no exception.
I think a model we’ll see for AI adoption and integration is one that we saw for digital and emerging technologies over the past decade. Large global companies will likely continue their pattern of creating centers of excellence for innovation tasked with researching, testing, and piloting projects related to emerging technologies in general. AI will be no exception – it will require education, investment, and the willingness to learn (and make mistakes) for companies to realize the full potential at the organization and brand level. Companies will also continue to expand the skill set of current and future hires. Just as companies across all industry sectors have expanded their head count related to data science, data analytics, and specialization in AI and machine learning, the pharmaceutical and medical device industry will do the same. Data scientists will be part of the conversation – working on cross-functional teams. Of course, these new hires, emerging technologies and investments in innovation must solve real business problems with increased efficiency and value if they are to survive internally – and not just layer new technology on an existing process
Any advice for companies seeking to better understand the impact of AI on their business, brands, and customer experience strategy?
In addition to companies experimenting with and building out AI-focused teams within a center of excellence, what we learned in the research we conducted with pharmaceutical executives was that they employ a range of build, buy, and contract. On the build front, some companies are successfully building out data science groups working across the organization. On the buy front, some companies are actually buying AI assets and startups – one prominent example being the acquisition of Flatiron Health by Roche. Last but not least, a large number of pharma companies rely on partnering and contracting with vertical AI solutions and technology platforms, as well as consultants with industry-specific experience. A number of large pharma companies are also tapping into the expertise of the large global technology companies – and platforms – defining and pushing the entire AI space forward. Those companies include household names such as Amazon, Google, and IBM Watson. All of these are examples of how pharmaceutical companies are learning by doing. Executives can attend various conferences on the topic of AI – but actually investing and experimenting must be a part of the mix.
As the relative role, and importance, of AI grows over time, who owns the technology or strategy within a pharmaceutical or medical device company?
That’s a great question. What we have learned in our research with pharma companies is that these AI teams – many of which are very early stage – tend to be cross-functional and work across various teams and parts of the organization. This is important because it shows how a technology such as AI and machine learning is not limited to one brand – or even one functional area. The core foundation of AI is often the data, so companies – and brands – must often work across the organization and with external partners to acquire, rent, or integrate key data sources used to make business decisions. Yes, traditional roles like technology, data management, and IT will continue to play a key role as companies invest and expand their commitment to AI and machine learning. However, brand, marketing, sales, and commercial leaders must have a seat at the table to ensure the tech is aligned with business goals, objectives, and key performance indicators (metrics).
Any examples of how AI is already impacting the business of pharm sales and marketing today?
All too often, we want a nice clean example of how one specific brand is using AI – such as a chatbot, an AI-powered treatment algorithm, or an AI-enhanced marketing dashboard. While these are great examples for teams – and companies – to discuss the practical use of AI and associated technologies, the bigger picture is where the overall action is happening today. If we step back for a second and think about healthcare delivery overall – AI is becoming part of the system at each step in the process. Providers are using AI platforms to incorporate patient data, therapy outcomes, and guidelines to make treatment decisions. The payer community is using AI to mine clinical and economic outcomes and determine how tech can improve and automate a process such as pre-authorization. In other words, in addition to the examples of a pharma brand using AI and machine learning to power a marketing or brand application, AI is also making a significant impact when it comes to how providers and payers are making decisions – today. In a world where AI powers clinical decisions by the provider and payer community, the rules of engagement – and influence – evolve. Pharma and medical device companies must become conversant with the technology, algorithms, and understand where they stand – and how to optimize the role of products within the new world of data-driven medicine.
What does modern pharmaceutical sales and marketing look like in 2025?
Most industry analysts can agree on one thing – the commercial model for the pharmaceutical industry will continue to undergo significant changes in the coming decade. We have already migrated from a world where the individual sales rep was the primary resource for educating physicians – and was a critical source of information to drive their individual decision making. It’s not to say the rep is not relevant in 2025 – they are – it’s just that the model and the pace of change accelerates even faster. Another area of rapid change is the use of data to optimize the customer and patient experience. Better use of internal data, and matching with third-party resources, will facilitate a highly customized user and customer experience. In much the same way that AI-driven data analytics will drive the R&D infrastructure and decision making of the future, AI and machine learning will drive the investments and the tactics employed for the customer experience – helping marketers focus their attention and budget where it has the greatest impact.
Longer term, when algorithms are making and driving clinical decisions – or greatly influencing those clinical decisions – the focus within the commercial model must evolve to incorporate data (and the use of that data) as a key part of the strategic sales and marketing mix. What algo is being used within a payer? What algo is being used within a specialty segment? What algo is being used by a government payer or regulatory body? In much the same way that an algo is used by Google to determine the relevance and placement of content within their search engine, the use of an algo (powered by AI) will determine access, coverage, and adoption rates for many therapeutic options in the coming decade. Does it make sense for pharma and medical device companies to better understand the data being used, to better understand what powers these algorithms? Of course. It’s going to be an iterative process – and will require collaboration across the stakeholders of the health system. Just as we learned about the impact of Google (and their influence) on providing access to content – there will be winners and there will be losers. Education, investment, and willingness to learn in these early stages will separate those leading the pack in the next era of data and AI-driven medicine and those struggling to adapt to the new world of healthcare delivery and reimbursement.
EVP, Intouch SolutionsConnect on LinkedIn
How is Intouch evolving to be the Agency of More for modern pharma marketing? How is AI part of those plans?
Every day, we are consulting with clients on modern pharma marketing – what it means, where they are, and where they need to go. We’re actively investigating how modern marketing and AI can help solve their business challenges, both short term and long term. This begins with assessing their readiness, but it’s already moved to developing and even deploying real solutions in-market like AI-driven chatbots powered by Cognitive Core™.
What’s Cognitive Core?
Cognitive Core is our proprietary AI platform. We built it to work specifically within the pharmaceutical industry and its regulations and restrictions. It ensures that the user experience is consistent, iterative, and personal, across multiple channels. Created with patient privacy and compliance in mind, it has adverse event and product-complaint recognition — along with notification and reporting abilities — built in. To accommodate a brand’s unique MLR requirements and associated process, Cognitive Core has a customizable workflow that can be implemented for each brand.
Is this all still theoretical?
Not at all! One demonstration of Cognitive Core in action is “Lea,” the chatbot we recently created for EYLEA.
We launched the EYLEA DTC campaign to help empower patients with wet age-related macular degeneration (Wet AMD) or diabetic macular edema (DME) to learn more about their disease, take an active role in its management, and have more meaningful conversations with their doctors about EYLEA.
There are three components of this multichannel campaign:
What is a chatbot for?
Chatbots are essentially virtual assistants that are featured on many websites to interact and assist the user in answering questions and provide additional content. “Lea” is built into the EYLEA patient website to answer questions and messages. She can keep patients or caregivers informed and answer questions on typically expected topics. If Lea does not have a pre-programmed answer to a question, she will try to redirect the user to content that may be relevant. Lea is not intended or positioned to be a replacement for the role of a healthcare professional.
What is the purpose of launching the chatbot on EYLEA.us?
Lea is a way to connect with patients and caregivers in real time to extend their learning. As a convenient means to access various educational topics and support along the treatment journey, Lea can help patients become better informed and more active participants in their eye health and care.
So, Lea is one example of what Intouch is doing with AI. What else can you tell us?
Intouch is shaping the future of AI in pharma as we continue to identify new use cases and create real solutions to validate their effectiveness. The use cases fall in areas of marketing, patient support, HCP engagement, and the creation of operational efficiencies.
Three common AI technologies are predictive analytics, robotic process automation and natural language processing. Intouch’s Cognitive Core blends all three to create valuable user experiences that can augment every aspect of a patient’s or professional’s journey.
Can you give any more examples?
Sure! The first one is a “banner bot” — a media ad powered by AI. It can solve a number of challenges, such as relevance, awareness, personalization, engagement, and value for an end user. A banner bot can be the entry point for a consumer into a user-centric, intelligent, and connected ecosystem that evolves with the consumer. Every interaction with the banner bot is an opportunity for the platform to provide additional value add to the consumer.
Banner bots can strategically engage users in additional channels like SMS and email and can even add users to a CRM stream. Cognitive Core seamlessly integrates the systems. By analyzing previous interactions in a secure and compliant manner, AI-powered banner bots can evolve the engagement experience beyond the traditional media banners.
A third example of AI power at work is in patient support: multichannel bot engagement with contextual conversation capabilities. What’s often missing in chatbot conversations is contextual relevance. This lack leads to forgettable conversational experiences. (We’ve all had them.) Contextual relevance is the essential bridge between technology and meaningful human interactions.
The first step in creating contextual relevance is an AI engine’s ability to create a dynamic patient/user profile that becomes smarter with every interaction.
Step two is for the AI engine to be able to look back at previous conversations to add a layer of personalization. Once the profile is created, it can grow over time, with gaps filled in along the way, enabling conversations to become richer and more personal.
The third step is the ability of the AI engine to connect patients with the right resources at the right times through the channels they access.
The fourth and the final step is to learn and recalibrate the previous three steps. So, why isn’t everyone doing this? Each step can be difficult to implement, especially in pharma. Technology isn’t the only challenge: there are issues with compliance, regulatory, privacy, and GDPR requirements. Cognitive Core’s conversation engine was built with contextual relevance and pharma requirements within its foundation. This makes it easier to build contextually relevant, as well as useful, conversations with consumers.
You have two more examples, right?
Yes! The next one is an example of AI improving HCP engagement, with an embedded virtual assistant that connects with Veeva and Salesforce.
Digital sales aids in pharma have become common. But even with their widespread adoption, many representatives still feel challenged by the amount of time and data it takes to research and document information on their customers.
The scope of activities required of a sales representative is vast. Most reps are required to see up to six customers every day, but the time they spend with those customers is surprisingly short. Instead, much of their time is spent planning the day, preparing for meetings, getting from place to place, documenting calls, and coordinating with partners and managers.
In the future, sales calls will be facilitated by AI, which can assist in planning, give helpful tips and do the necessary but laborious follow-up tasks.
Companies need an efficient way to identify and communicate HCP needs and requests to reps in preparation for their calls, so that the most applicable and available content can be shared during sales appointments. AI can automate tasks, give proactive reminders, and create reports and analytics automatically, among other tasks. Cognitive Core integrates with DSA platforms like Veeva, assistants like Alexa, and tools like SMS to completely re-envision for reps how a typical HCP call occurs. This includes pre-call planning research and insights, AI-based insights just before the call, and post-call activities such as logging a call.
In Chapter 1, we learned how the proliferation of data is changing the world, as digital data becomes ever more integral to our daily lives. In Chapter 2, we had a bit of “AI 101” as we learned what some of the most common types of AI computing are and how they work.
In Chapters 3 and 4, we got personal, learning about the exciting work that the top players in healthcare AI are doing today. Then, we discovered the opinions of pharma marketers on how AI is affecting — and will affect — modern marketing.
But now, we look ahead. From 2018 to 2023: How will AI cause pharma marketing to evolve over the next five years?
As we said in the Introduction, "we see it as our responsibility to help discern inflated hype from real hope.” It’s important when making decisions every day, but it’s even more important when we’re trying to look ahead into the future. It’s very easy to take a small positive or negative event and over-extrapolate its importance.
One useful tool for discerning hype from substance is the Gartner Hype Cycle, an industry standby, which helps to view innovations and the inevitable swings of overpromotion and disappointment that eventually level off to productive use. View the video below to learn more about the Gartner Hype Cycle and where AI currently fits into the curve.
As with every other technology, AI has limitations. It will neither create nor solve all of our problems. But there’s no denying at this stage that healthcare is an area where AI shows some of the greatest promise. As McKinsey Global Institute leaders note in this recent podcast episode, McKinsey Podcast: The Real-World Potential and Limitations of Artificial Intelligence – healthcare is likely to be one of the industries where AI has the greatest financial impact, at nearly $400 billion.
None of this is imminent. Only about a quarter of pharma marketers in our survey are currently using AI solutions in their work. However, more than 40% have a team dedicated to AI strategy and/or implementation.
We believe that AI today is analogous to the “e” revolution of the 1990s. By 2023, we won’t be talking about AI-powered solutions – the same way that we eventually stopped talking about “e-business.” Terms that specify a technology’s inclusion stop being used when the technology becomes an assumed part of life.
So: must you run out and create an AI project today to keep up appearances? Absolutely not. But AI is going to become a fundamental part of modern pharma marketing in fairly short order. This will happen in both blatant and subtle ways – from specific AI tools, to behind-the-scenes use of AI technology. And it will happen whether you prepare now or not. We’re here because we’re fans of preparation.
We’ll learn how to better deal with data – because we’ll have to. How to store and manipulate enormous quantities of information. How to keep it safe and secure. Who has rights to own it, to access it, to control it, to profit from it. This will touch every industry. And the details will be battled in courts and debated regulatory and standards bodies for years.
As AI continues to become more common, we will also see “false AI” continue to proliferate: standard computer coding that looks a lot “smarter” than it is. This isn’t necessarily a bad thing (unless, of course, it’s being sold as more than it is). There’s going to be a place in the world for standard computer coding for a long, long time.
We will slowly notice the technology in our lives being able to assist us more: to give us better predictions, to be able to do more with less instruction. Phones will get even smarter. Language processing will become sharper. Predictions will be more accurate and helpful. The idea of a “service robot” sounds hopelessly futuristic. But what is a Roomba, or a self-parking car, but a service robot?
As we noted in Chapter 2, “What Is AI and Why Does It Matter,” AI allows for stronger, more adaptive, directed marketing, more efficiently, more cost-effectively, and with more assurance of compliance. That’s a combination that’s hard to beat.
Industry analysts agree, pointing out how AI’s ability to understand data is particularly suited to benefit marketers.
Chatbots triage customer service requests. Hyper-smart marketing automation adjusts itself for improved ROI. Predictive models help us to better understand adherence and drop-off risks. Many AI-powered applications like these already exist. They will strengthen in number and in power over the next five years, and will be joined by others that stand on their shoulders to go even further, particularly into coverage, reimbursement, and treatment decisions.
Even if you’re not using AI to execute marketing at the brand level, you must understand how data is being used to make clinical decisions. How do you prepare, and respond, if and when an algorithm becomes the key influencer? How will you market for that, educate for that, influence that?
In Chapter 1, we demonstrated how NVidia used a neural network to mimic a world-class composer. Similarly, AI and modern marketing will enable marketing “conductors,” providing them with automation and insight that not only free them up to focus on strategy, but give that strategy power and assurance.
In many ways, the role of the advertiser has already shifted. Marketing was once largely an art form in which creativity reigned, but today, data often beats creative. Amazon’s marketing success, for instance, is driven not by complex creative, but by putting enormous quantities of data to use for predictive personalization. Creativity will never be unimportant; but rather than performing solo, it is now part of a more successful ensemble.
A recent survey by the World Federation of Advertisers found that the greatest skills gap between current and future needs is AI predictive modeling. Simply put, the modern pharma marketers who succeed will be the ones who understand what successful people have always grasped: what tools to work with, how to use them, and how to work with the right people.
Just as the digital revolution of the 1990s required individuals across the company to build their knowledge and work together, the AI revolution will do the same. Teams across organizations will be tasked with working together to find new and better ways to identify, collect, share, analyze, interpret, and act on data.
Amara’s Law says that “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” Five years, in many senses, isn’t that far away. But for technology, it can be several lifetimes. In 2023, we’ll see AI in healthcare in ways we’re not even considering right now.
Perhaps we’ll be able to monitor the impact our every daily decision will have on the future of our health. You might be less likely to skip flossing today if you’re notified that doing so will cause a 2% jump in the likelihood of a 2030 root canal.
While that’s happening, perhaps the nanobots in your bloodstream will be rooting out an errant cell they’re able to predict would have become cancerous.
But whatever the specific achievements of even the greatest health AI solution are, it’s likely that the tectonic shifts to the industry will matter even more. Done right, healthcare in AI will help to improve what is now an immensely complex, expensive, emotionally laden field full of questions and imperfect solutions.
None of us can see the future (which is part of what makes it so exciting). But all of us can prepare for it. Both the Digital Health Coalition and Intouch Solutions are incorporating AI into our work today and our plans for the future.
The Digital Health Coalition and Intouch Solutions work to remain at the vanguard of technology-driven change in life-science marketing.
In “Modern Marketing: Pharma's Data-Powered AI Revolution,” we’ve addressed how the proliferation of data is causing radical shifts in marketing; what AI is, how it works, and why it matters; and what the current and future states of modern marketing look like.
Our current time is an exciting one. The years ahead will require much work and innovation, but will deliver immense results to those willing to invest the time and effort. This is the perfect time to evaluate next-generation tools and begin to learn how they fit into the toolbox, and mindset, of modern pharma marketers.
We welcome your feedback and comments. Was this ebook helpful? Is there other content you’d like to see in the future? Did we leave questions unanswered?