At the 2019 Pharmaceutical Marketing Science Association Conference in San Diego over 300 pharma analytic professionals attended multiple presentations on the conference theme of “How to Design the Right Analytics Campaign that will Reach Your Targeted Market.” The most relevant constant across all of those presentations was the need for a data scientist on your commercial analytics team. The ideal data scientist has an advanced degree in mathematics or statistics, has experience working with databases and can program, and has in-depth domain knowledge of the pharmaceutical industry. Think that person is hard to find? You are absolutely right. That is why most big pharma is setting up a Data Science Team and not focusing on one guru to do it all.
Data Engineer
Vitaly Gordon, Head of Engineering at SalesforceIQ says, “You don’t solve your data problem by choosing the right technology.” As good as all the software is out there, having a strategy for utilizing it the most effectively is critical to getting going in data science. Gordon goes on to say your first hire for your Data Science Team has to be a Data Engineer. It is possible your IT department will have someone with that expertise but without some knowledge of statistical applications the data won’t be set up in a manner that is useful down the road. Legacy systems and multiple databases with multiple sources of information have to be integrated to build an effective infrastructure for data analysis.
Data Analyst
Your second hire for your Data Science Team might be a Data Analyst who can maintain and clean that database. Garbage in means garbage out is always true of data. Both of these positions, Data Engineer and Data Analyst, are focused on output that is internal in nature. The external work, the building of a product to improve some end user experience, is the work of the Data Scientist.
Data Scientist
The challenge in hiring a Data Scientist in the pharma industry is getting someone with domain experience. The best data scientists come from industries further along in the process of utilizing their data: technology, financial services and vendor organizations. Insurance has statistical sophistication from the data processing side as well as their actuarial groups, but even the degreed statisticians on the actuarial team are just starting to use data science effectively. Health insurance statisticians would have more experience integrating data across the myriad of healthcare platforms, patient records, pharmacy records, clinics, lab results, and more. True pharma experience will be a rare find.
Vitaly Gordon from SalesforceIQ goes on to say that the premise, “if we just flood people with information, they will know what to do” is not the way to approach data science. Those implementing a Data Science Team in the pharma organization need to spend time BEFORE hiring on determining what the goals are of the analytical program. Where do we want to go and what is the value?