Understanding customer behavior and making strategic decisions utilizing 'Big Data'

Every industry needs more information about its customers—how to find them, how to reach them, and how to sell to them. In nearly every industry, the customer is direct, whereas in the pharmaceutical industry, a customer can be one of many indirect entities. Historically, pharmaceutical companies have seen their primary customers as the healthcare professional or prescriber.

But the definition and scope of a customer has always been a complex equation in this industry. Ultimately, it is the patient, of course, but what if the historic ways of reaching the patient are all wrong? What if the real customer is a referral network or insurance payer? How will that change the way a pharmaceutical company develops and describes a product? And how will a company identify its customers and how to reach them?

The era of Big Data is raising a lot of these types of questions, and now, for the first time, pharmaceutical firms can answer them. This new era in the pharmaceutical industry is largely the result of new federal regulations and the advances of tools for analyzing huge troves of data. New laws, such as the Patient Protection and Affordable Care Act and the Sunshine Act, have forced manufacturers to develop cross-functional systems and internal processes to capture and aggregate data.

The rise of “Big Data” has been described by prominent computer science professor Joe Hellerstein of UC-Berkeley as “the industrial revolution of data.” Craig Mundie, the head of research and strategy at Microsoft, correctly summed up the status of the “revolution,” when he said, “You can see the outlines of it, but the technical, infrastructural and even business-model implications are not well understood right now.”

The sheer amount of data that is already gathered every day (and will increase because of the Sunshine Act) has the potential to provide insight into many areas of healthcare. From marketing channels to physician presentation impact information, the data are simply out there to help optimize commercial efforts.

All of the targets of pharmaceutical companies form a complex healthcare ecosystem that is crucial to understand if firms are to successfully identify customers. But to successfully use these data, pharmaceutical companies will need to be integrated. The traditional pharmaceutical enterprise is generally comprised of “feudal data lords” that acquire, create and maintain their own data systems within their own functional areas; there are many barriers for the sales and marketing organizations to utilize data from the Key Opinion Leader (KOL) or Clinical Trial Activity Systems. The era of Big Data allows manufacturers to join these data silos together. The amount of data that will be collected will increase the current data by at least a factor of 10, but armed with that new data, pharmaceutical companies will acquire for the first time, a true 360-degree view of their customers.

One warning: Much of the data will come from aggregate-spend and disclosure regulatory filings, so pharmaceutical manufacturers will need to be careful to use the data in a compliant manner. Is that possible? Absolutely. For example, a manufacturer could combine historical data sets (prescription records, sales representative activity, data from managed care contracts, etc.) with new data sets (aggregate spend activity, social media verbatims, etc.) to create maps of “influence networks” that help pharmaceutical organizations determine true KOL profiles. Fig. 1 is an example of algorithmically determining influence networks; Fig. 2 provides a snapshot of how to simplify and integrate these new insights with existing KOL and commercial tools.

Collecting data in a compliant manner

The rapid increase in the capabilities of data capture and analysis technologies has made it easier to gather and analyze previously unavailable healthcare data, while remaining legally compliant. Never before has there been so much information available to draw upon to understand customer behavior.

The retail industry was the first to tap this firehose of data and to understand the new social media promotional channels available. If you have ever shopped at Amazon.com, its logarithms likely identified what you were going to buy before you did. While the pharmaceutical industry can learn from the retail industry, there are certain unique challenges to the industry that must first be overcome. Pharmaceutical companies must be sure to understand how to use healthcare data and how they can be used while remaining compliant with the law.

There are four major concerns with using healthcare data for sales and marketing:

• Impact of Sunshine Act regulations on commercial activity
• HIPAA concerns
• Compliance controls over enterprise data
• Cross-channel challenges

Sunshine Act

The Sunshine Act was created to increase transparency and accountability in healthcare payments by requiring the collection and reporting of physician spend data. For a number of years, pharmaceutical companies have been collecting spend data on a state level. As a result, a number of healthcare providers have opted to no longer accept any industry spending. With the passing of the Sunshine Act as part of PPACA, the regulations may impact current promotional programs, such as speaker programs and ad boards, and the loss of KOLs for such critical activities as clinical trials and consulting. If they were not accepting or reporting any industry spending, this also represented a loss of data that was helpful to analyzing the industry.

HIPAA concerns

However, the data from new, additional sources contain great potential for marketplace analysis. One of the many functions of HIPAA is to provide security and privacy regulations around health data. Much of the potentially available data currently housed in hospital systems was gathered through the use of Electronic Health Record (EHR) software. That is a potential treasure trove of data from healthcare technology vendors, such as Epic, Cerner and Allscripts, as long as HIPAA privacy regulations are respected.

Cross-channel challenges

While the Sunshine Act and HIPAA concern federal laws, the cross-channel challenges address the fact that many types of data (i.e., prescription data, Personal Health Information) cannot legally cross channels without violating patient privacy rights. There are several ways to collect and analyze healthcare data today that allow the data to be used legally. To properly screen and use this information, pharmaceutical firms will need to create policies that assure information security and the quality of data. Subsequently, a multidisciplinary team—potentially including representatives from the legal, public relations and information technology departments, and data scientists—should be created and charged with addressing and determining the role of Big Data. Training programs must be defined and implemented on the usage of data by cross-functional teams. Concurrently, strong business rules must be defined and implemented for data that cannot cross paths.

Big Data technologies, such as Hadoop and MongoDB, are emerging to help optimize the collection and processing of the data deluge. Unstructured or NoSQL database technology is helping break down the barriers that currently exist in collecting heterogeneous data at a large scale.

Developing operational decisions

Once data-set modeling is deemed compliant, analyzing data and understanding customer behavior is the key to help with the strategic decisions of a successful business. New visualization technologies, such as Tableau, TouchGraph Navigator and NodeXL, are emerging to help understand these patterns of behavior, given large amounts of data. These tools will also help identify the various interdependencies. For example, the ability to identify and track various peer networks of HCPs can be visualized. The data can determine the number and the specific levels of HCP touch points an individual can have. These connections are the foundation to determine the influence of the network, as well as the comparative value between networks.

The vast amount of data available on the Internet in blogs and social media contains useful data on the behaviors of KOLs, but they must first be identified. Once identified, patterns of behavior can be mapped for these KOLs, and appropriate actions can be taken. One key benefit of Big Data technology is the ability to consume unstructured data rapidly and perform analysis that would have taken months previously. This gives the data scientist the ability to have “more at bats,” with the prospect of getting more hits. Big Data clustering technologies, such as Cloudera and their packaging of the Hadoop stack, are crucial in processing the high volume and high velocity of these new data sets.

Big Data can also help pharmaceutical companies understand their own operations. For example, aggregate spend data can be combined with census data to determine where the highest and lowest spending is occurring. Companies can discover spending categories that must be reined in, and identify potential compliance issues.

Sales, marketing, operations and compliance professionals are about to access a gold mine of information about how their products are being talked about and used. The data are there for the taking, and with care, it can be gathered and analyzed without violating privacy restrictions. That data can help companies identify trends, such as patterns of customer behavior, and determine strategic direction, often in real time. The market is just becoming aware of the possible uses of this new information. Wise, resourceful companies will want to start to understand its usefulness and create a strategy to employ it as soon as possible.

ABOUT THE AUTHORS

Manny Tzavlakis is a managing director at Huron Life Sciences and is based out of the New York office. He has 17 years of pharmaceutical industry and management consulting experience specializing in aggregate spend/disclosure, predictive modeling/data analytics and various commercial operations areas. Manny previously worked for Fair Isaac Corporation, Dendrite International, Accenture and Novartis Pharmaceuticals Corporation.

John Poulin is a director at Huron Life Sciences and is based out of the New York office. He has spent more than 15 years working in the pharmaceutical, healthcare, information technology, and consulting industries. He specializes in developing information technology systems to enhance business process, meet regulatory requirements, and mitigate risk.

Sam Chud is senior director of business operations at Shire Pharmaceuticals. He has more than 20 years of experience in the pharmaceutical industry and is the former director of salesforce effectiveness at IMS Health. Earlier in his career, he worked for Bristol-Myers Squibb and Bracco Diagnostics.