Highly complex modern IT organizations across all industries have been tasked with the challenge of not only keeping up with the rapid pace of technology, but also deploying it throughout their companies in an efficient and secure manner. In doing so, the “I” in “IT” sometimes gets overlooked or at the very least ,it can take a back seat. But not for much longer. With the acceleration of big data analytics, I believe information, and therefore IT, are set to experience a major renaissance. The trick is making sure we go about it in the right way.
The merging of Health and Data Science
When I began my role as Chief Information Officer at Merck, health data was expanding at breakneck speeds with the proliferation of electronic medical records, the growing ubiquity of biosensors, and the affordability of genome sequencing. Data scientists could detect depression based on an individual’s changes in movement and phone call patterns. They could predict illness the day before onset due to mobile devices sensing less physical movement of their owners. I was witnessing data-savvy companies from outside the healthcare industry catching onto this huge data goldmine quickly.
At Merck, we had just successfully implemented critical global information systems; however, we still had work to do to further exploit that data for business gains. In my mind, it was time for us to think about ourselves as a data-driven company and become an analytics powerhouse. After all, the lives of our current and future patients could benefit greatly from it.
Whose Job is it Anyway?
But there remained many unresolved questions. Whose job is it to bring all of the internal and external data together at scale? Who will be the chief advocate for big data analytics across a multi-layered enterprise? And who will hire, develop, and inspire the teams to create the advanced tools and algorithms that could unlock the power of data? For me, it was a no-brainer. There is no better function for the job than IT.
Why? First, IT is one of the few functions in an organization that truly has the end-to-end visibility across the enterprise. That visibility allows us to not only reach every part of the organization readily, but it will also help to ensure that we are capturing the full value of our data by leveraging it across organizational boundaries. Second, IT has the great access to the data that exists across the company, which is an essential asset when building data warehouses. Third, we know how to get the most out of our information through the use of technology, which certainly doesn’t hurt in this space.
These characteristics on their own don’t ensure success. For example–we needed some boundaries. Analytics could never, and should never, be an IT-owned initiative. That would all but guarantee our failure. For Merck to really capture the true power of analytics, it must be owned by the entire company and catalyzed, empowered by IT. So that’s what we set out to do.
The Business Analytics Practice– Empowered by IT
As a pharmaceutical company, Merck consumes scientific data to advance our medicines as well as business data to drive operations. Since Merck’s R&D organization has been successfully embedding scientific analytics into its standard practices for many years, we chose to focus first on business performance analytics and established a cross-functional Business Analytics Practice. The team is comprised of individuals representing every region and divisions across the company, and through it we determined three key areas, or stacks, where IT will play a pivotal role.
Stack #1: Data Liberation & Integration
Every time an employee logs onto the network, a customer or patient calls into a call center, someone clicks into the website or a product enters or leaves inventory–data is created. In the past, this data sat in silos, either inaccessible across the company or simply not speaking in the same ‘language’ from division to division or region to region.
IT, with its end-to-end visibility across the enterprise, is the ideal owner of the data integration process and the best function to liberate it through a data store so it is easily accessible to anyone in the organization.
Stack #2: Analytics Libraries, Models, Algorithms, Indexes, and Self-service
With all of the data integrated and accessible in a big data store, we needed a way for business colleagues–not all of whom have a data science background–to access it, read it and analyze it in a way that could lead to valuable insights.
In this case, IT’s role is to not only provide the models and algorithms that will turn raw data into meaningful information, but also to do it in a way that is easy to use, digestible and repeatable for the everyday user. Facebook, for example, calls this the democratization of data–making it easily consumable so that it can be provided in a self-service way to a diverse group of teams who may or may not have any data background.
Stack #3: Analytics Catalyst
It doesn’t matter how integrated and liberated your data is, or how advanced your algorithms are–if you don’t have business managers who are aware of (and open to) data-driven decision making through analytics, the program will fail.
At the end of the day, our business colleagues own the outcome–not IT. However, IT can act as a catalyst to help teams not only understand what analytics can do for them, but also guide them to ask the big tough questions that will allow data analytics to perform its ‘magic’ and achieve the highest impact. Delivering Value Merck’s Business Analytics Practice realized more than $700 million in value in 2014, but that’s only the beginning. We are getting smarter and more efficient as we partner with various groups across the organization. In fact, IT has a better understanding of business drivers than at any point in our recent history, affording us a seat at the table for many strategic business discussions.
Most of all, though, our work with our colleagues is empowering Merck with the information and data-driven decisions that could save and improve more patients’ lives around the world.