25 January 2019
Leaders at Novartis share how the big pharma is laying the foundations for a potential new operating model as a "medicines and data science company"
At the beginning of 2018, Novartis CEO Vas Narasimhan spoke of reimagining Novartis as a medicines and data science company. The drugmaker was sitting on a “goldmine” of vast datasets, “built from having conducted countless studies in thousands of diseases.” Taking advantage of this “wealth of clean, curated, longitudinal, and interventional data,” Narasimhan explained, “has the potential to completely transform the way we develop medicines.”
Novartis had already begun signaling its commitment to the plan with a series of high-profile appointments of data science experts from outside the industry. In August 2017, the company hired Bertrand Bodson to the new role of chief digital officer. Bodson had no pharma background; he had previously served as chief digital and marketing officer for the UK retailer Sainsbury’s Argos, where he played a pivotal role in guiding the company’s digital transformation from aMiriam DonaldsonMimi Huizinga, MD
catalogue business “into a multichannel online powerhouse with more than a billion visits online annually.” Earlier, he had spent three years as Amazon’s senior group product manager. “The pharma space in particular is ripe for disruption,” Bodson commented in October. “Looking forward, data science will drive more of our decisions and more of our programs—including our medicines—will be digital.”
Following more appointments of data experts from outside the industry in 2018, Pharm Exectalked to two key players in Novartis’s push for digital transformation—Mimi Huizinga, vice president of strategic data and digital, and Miriam Donaldson, head of HR Digital—about how this influx of new talent is fueling a cultural change that is set to define the future of the drugmaker’s data science activities.
PE: Can you outline Novartis’s mission to transform itself into a medicines and data science company?
MIMI HUIZINGA: Novartis associates are learning to be data curious and more agile in how they approach their work. We are seeking ways to use data and digital to support every aspect of the company—drug discovery, trial planning, manufacturing, commercial operations, talent management, and many more. It is truly a transformative time.
MIRIAM DONALDSON: Through our digital transformation, we aim to be the leading medicine company powered by data and digital. If you look at external research about what leads to successful digital transformations, a common theme is capability building. This is a priority for us as well. We are looking at capabilities across three different populations at Novartis: all associates, our leaders, and our digital practitioners. We are building different capability solutions for each, as they have different needs from general awareness and understanding of the potential of digital to leading teams who feel empowered to experiment, to having world-class data and digital expertise.
Importantly, we want to encourage all of our associates to be curious, and feel comfortable around data and digital, and help us to see how these things can work better for us. We need to set them up for success to be part of that journey, so investing in their digital awareness and capabilities is really critical.
PE: What are the challenges in achieving this transformation? And what advantages do you have?
HUIZINGA: We want data and insights to drive our digital solutions, but healthcare data is difficult—the data is siloed with no single source of a complete, longitudinal record. This means that we have to deeply understand the data and its limitations to ensure we use the right dataset for the question at hand—and that our analytics help to minimize the risk of false conclusions from the inherent biases in the data. We are not able to understand patient barriers in the care journey as rapidly as we would like and that slows our development of potential digital solutions.
DONALDSON:To lead a digital transformation, you have to be comfortable with failure. You have to be excited to experiment, learn from what works and what doesn’t and then act on it. You have to do it quickly and build solutions that you can scale. But that’s not the culture we’re starting from at Novartis. We have to help shape an environment where curiosity can really thrive and we have leaders who support their teams, not by telling them what to do but by creating a space where they can experiment and learn and not be frustrated if things don’t work the first time. Novartis is also transforming from being a collection of seven interdependent companies into one integrated company. This idea of building solutions that scale across Novartis is new for us. There’s a cultural aspect to this as well, we’re moving from rewarding individual heroes to teams that systemically solve problems and design innovations that scale.
HUIZINGA: Novartis does have many advantages—our mission, our people, and our commitment to data and digital. As an MD and former epidemiologist, I joined Novartis because of the core focus on patients. That focus really clarifies how we make decisions and set priorities. There is a fantastic set of data, analytics, and digital experts within Novartis. We are working now to create more connectivity across the company. Finally, having leaders that understand the importance of data, analytics, and digital makes a tremendous difference.
PE: You’ve made key new hires from companies such as Google and Amazon. Can you talk about the importance of hiring from other industries?
DONALDSON: We approached this as choosing people with the portfolio of skills and experiences that will be important for the transformation. It started with the hiring of Bertrand Bodson as our first chief digital officer, who came from Amazon and retail, and had no experience in pharma prior to joining Novartis. For other key roles, such as our head of data strategy and head of data science, it was important to us that the folks we hired at this level had experience leading transformations.
That’s what led us to Raj Patil (head of data strategy) and Shahram Ebadollahi (head of data science and advanced analytics). Raj, for example, worked in the financial industry, which is also highly regulated and has some of the same challenges. He had successfully built data strategies in several banks and also worked at Google. Shahram helped to build the IBM Watson Health organization, a 4,000-person team. Through his work there, he experienced many of the use cases we are looking to address and has learned what works and what doesn’t. We believe that pairing that experience with our incredible in-house team of healthcare experts is going to be a powerful combination. Possibly most important with both of these leaders is that they have leadership styles consistent with the culture we aspire to and which we feel will engage our teams to work together in new ways to innovate in data and data science in close partnership with the business.
With the new talent we’re acquiring, especially in the data science area, the questions we ask ourselves when they come into the company are: How do we set them up for success? How do we help them to understand our organization while also asking them to help evolve it? We’re not trying to build a separate digital business, we want to transform our business using data and digital, which means integrating and influencing both what we are doing and how we do it. Specifically, we’re focusing now on how we elevate data science at Novartis, which in the past has maybe been viewed more as a supporting function to our clinicians, and other types of scientists and reposition the organization as a true strategic partner to the business.
PE: What specific initiatives and policies are you implementing to drive technological and cultural change?
HUIZINGA: Broadly speaking, our activities fall into three buckets: data, analytics, and impact. For data, we are seeking to optimize the use of internal and external data while looking for new types of data that might provide value. We are working to organize the data and enable access to all appropriate employees while keeping data privacy top of mind. In the analytics bucket, we are automating our core analytics so that we can free up our analysts to think about the business issues and more complex analysis. We are also using artificial intelligence and learning how to industrialize AI models in our daily work. In order to see impact, we use design thinking concepts to ensure that results are meaningful and that visualization tools are intuitive. We also look to see how we can automate pull-through and build digital solutions that fit within an associate’s workflow and support the patient journey.
I am working to promote a data curious culture in our US Oncology group. I work in all of the areas—data, analytics, and impact. I love seeing the “ah-ha” moments and hearing about how we helped someone think differently about a potential opportunity. My job is to ensure that when we build the data infrastructure or the analytics framework, that it will be able to meet our evolving needs three to five years from now. To do that, I have to understand our business strategy and be able to translate the commercial and medical leaders’ needs to our IT and analytics colleagues. On any given day, I may have conversations about our metadata management plan, detailed review of a Markov model we are bringing into production, brand level launch strategy, real-world evidence needs for a cost-effectiveness model, clinical trial feasibility, ideas for our next visualization product, and our website content management strategy. All of this makes my job challenging but very fun.
DONALDSON: One of the things we’re doing in our capability-building is around teaching our leaders how to help teams who need to work differently, such as agile working, which has been really important in software development in other tech companies. It’s a completely different way of organizing teams and pursuing the work that needs to be done. This is one of the ways that we’re trying to get teams more comfortable with the rapid innovation cycles, customer-centric solutions, and building solutions we can scale.
For things this agile to work at Novartis, our leaders need to lead differently. We’re building a program that we call Digital Immersion for Leaders, which is a three to four-hour experience for teams. It’s kind of a “choose your own adventure,” where we build cases around different scenarios at Novartis. We show leaders how digital is already changing the work we are doing and provide them with some leadership challenges, so that they see the impact that they have on teams who are trying to innovate, and work in the way that we aspire to—and all while seeing how these solutions address their real business challenges. We launched a beta version of this program to our top 300 leaders back in September, and we will be rolling it out to the larger company over the next several months.
PE: How will the next couple of years play out in terms of the transformation?
DONALDSON: Now that we have our heads of data strategy and data science onboard, we are ramping things up quickly. So, I expect in 2019 that we’re going to be making a lot of hires in that space. And as our associates, leaders, and practitioners benefit from our investments in their capabilities and our culture and ways of working evolve, we think it will feel really exciting to work at Novartis. Almost every element of our business can be touched by this transformation, so that we can make smarter decisions based on data that help us bring meaningful medicines to patients faster.
HUIZINGA: I have been amazed at how much we have accomplished since I joined in August 2017, but that work was largely foundational. We have really gained momentum over the last three months. We have a few tools deployed to the business and great engagement of our leaders. We are planning to launch an internal visualization tool [this month] that will provide deep insights to our commercial leaders. Our goal is that these tools will help the leaders ask better questions faster.
We have industrialized one AI model and over the first half of 2019, we will be testing the limits of that model to understand when we need to refresh and rebuild. We will also place the outputs of a second AI model into pilot mode during that time.
By mid-2020, our goal is that everyone in US Oncology will have a unique visualization product that supports their work and that stems from a common set of data and analytics.Print
08 July 2020
08 July 2020
07 July 2020
07 July 2020