The pharma industry now has an incredible amount of data at its fingertips thanks to electronic medical records, social media, wearables, and more. But just having access to this data isn’t enough—marketers need a way to turn all of this unstructured information into actionable insights. Not only that, but they need to do it as quickly as possible so they can act on these insights and in way that doesn’t require too much help from IT. PM360 asked 12 experts in the fields of data and analytics:
- How do marketers determine what information is most important or can be the most beneficial? What data sources should marketers focus on?
- What should marketers look for in a data analytics platform?
- What is best way marketers can leverage big data to have the most impact on their brand?
- How has access to big data affected marketing mix modeling? What is the best way to ensure you are optimizing your marketing spend and achieving the best ROI?
- How can big data help pharma improve its relationships with customers?
Analytics has become a household term. The trend is readily observable in the staggering investments of more than $50 billion per year in Silicon Valley, reminiscent of the pre-Y2K dot-com rush. Not unsurprisingly, it has also become a term that is oft misused. An entry in a software marketing literature that claims “Analytics” as one of its features gives it a rite of passage into this lucrative market, whether or not it truly provides a value beyond the quintessential number crunching features of Excel.
Finding insights is not trivial. Dashboards are good only to the extent that the end-user can derive information that has practical business implications. It is easy to get sidetracked by the marketing pitch, to believe in what we are told without the facts being substantiated with first-hand empirical results. The capabilities of a solution that works for certain organizations might not generalize to the needs of one’s own organization.
Developing a Successful Analytics Platform
To that end, a rigorous proof-of-concept—a complete assessment of current needs and future deliverables and their true cost to the organization, including resource, licensing, hardware, and other costs—is vital to the success of any analytics initiative. An enterprise-wide data analytics platform is usually a combination of platforms from different vendors that work in tandem and deliver capabilities to efficiently analyze vast amounts of information, run algorithms using open-source solutions, connect seamlessly with popular visualization solutions such as Tableau, and scale as the needs of the organization changes. The right tools in the hands of a skilled professional can indeed make a difference, and hence, investing time and effort in building a well-crafted future-proof, scalable, and adaptable platform is perhaps the most important aspect of creating a successful enterprise analytics platform.
Our technology-driven world is dramatically changing our lives, particularly as the pace of digital transformation is empowering us to act as both patients and consumers of healthcare information. As part of this new “Information Generation” (http://bit.ly/2bca1Xt), we’re always connected, using many devices, and generating data anywhere, anytime.
Smart technologies, like wearable devices, are driving the “Quantified Self Movement,” and arming patients with personalized big data to make more informed decisions about their lifestyles and care. These sensors can help remotely monitor physiological measures outside of the acute care setting, drive prescription changes by physicians based on real-time trends, and ensure proper medication adherence. With these “smart” patients, expectations for healthcare are shifting from a focus on one patient care episode to an overall patient lifecycle.
What to do With Patient-Generated Data
Healthcare leaders are already incorporating this growing patient-generated data into their ecosystem to move forward accountable care initiatives, as HHS Medicare payments are shifting from volume-based to value-based reimbursements, moving to 30% by the end of 2016 and 50% by the end of 2018. In tandem, pharma marketers need to accelerate their multi-tiered engagement strategies—continuing education to caregivers and payers on drug efficacy balanced with cost, but also creating more targeted and differentiated brand value messaging to the “empowered patient” for preventative health and disease management.
But, how do pharma companies leverage this next wave of data? Through big data analytics and the data lake. By creating an ecosystem of customer data centered on the patient and outcomes, pharma companies can gain additional insights on disease management and medication decisions as they justify new investments. Moving to a modern data analytics strategy that considers how to best incorporate customer-generated data will provide competitive differentiation in the marketplace and ensure pharma marketing campaigns have an immediate, personalized impact.
A look at the big data landscape highlights the fact that its full potential is not realized by pulling just one technology lever, instead a framework of technologies, data scientists, and systems must be created to capture quality data and synthesize it. With that in mind, pharma marketers should be focused on three opportunities presented through big data.
1. Small Data
One of big data’s biggest opportunities lies in “small data.” Small data includes the insights that can be observed throughout the customer experience (CX). Diagnosing the friction points along the health experience will inform moments of brand relevance, and this is how brands can better create consumer value.
2. Artificial Intelligence
A second opportunity for big data and brands is Artificial Intelligence (AI). Even though AI/deep learning has been around for a while, only recently could it be applied to huge sets of data economically and swiftly. AI is now helping big data live up to its full potential by assisting data scientists as they create predictive insights and models to inform brand optimization.
3. Data-Driven Ideas
Ultimately, for brands it’s about actionable insights that inform data-driven ideas. This focus on continuous, iterative improvement based on real-world performance allows brands and agencies to be more responsive to changing user needs, hard data from asset deployment, and changing conditions in the marketplace. By designing campaigns that can continuously leverage data and feedback to evolve, brands become more focused and relevant. And this is the impact of big data.
Big data arises naturally from the myriad interactions we conduct with organizations, individuals, and even appliances—captured as a digital resource for advanced data analytics. In healthcare, we tend to think of ourselves as awash in it, and struggle to tame the datasets that promise the greatest commercial insights into key stakeholders and market dynamics.
Rather than focusing on managing messy “naturally occurring” big data, however, we might be better off thinking more deliberately about creating big data that serves our needs right from the start—targeted and clean data, engineered to address critical marketing issues of interest, and ready for straight-forward statistical analysis. Today, the technology exists to do just this—by placing stakeholders, be they healthcare professionals, payers, or patients, into digital simulations that allow us to capture their behavior in a virtual environment.
Insights Through Simulations
We are not talking about immersive “virtual/augmented reality” as in The Matrix or Pokemon Go, but rather powerful digital simulations that engage stakeholders in healthcare-related tasks via their computers or mobile devices. Everything stakeholders do in these simulations is recorded, generating large streams of data related to information accessed and actions taken. By leveraging principles of experimental design in the creation of these simulations, we can ensure that observations of virtual behavior yield insights not only into what is done, but why—all without asking a single question.
Simulations like this are beginning to be used by healthcare market researchers—most commonly, for physicians to treat virtual patients under a variety of environmental conditions related to access, indication, guidelines, etc. But the utility of simulation-generated big data goes well beyond market research. Significant applications exist within health economics, sales, and strategic planning and should be on the radar screens of pharma’s commercial leadership.
When deciding on a data analytics platform, companies must consider several factors. For any data analytics platform to be successful, it must be designed and tuned by a team that understands the patient journey, decision processes, and the limitations of the data. With this foundation in mind, companies should focus on key elements for the tool, including the ability to enable quick customization, automate models to achieve scale, and create charts/graphs that communicate results effectively to a non-analytics audience. The more automated the platform, the more essential it is to also enable human input to separate signals from noise.
Using Data to Optimize the Marketing Mix
Today, digital and mobile channels for marketing can be optimized near-real time and specific audiences can bid at different levels to maximize returns. However, some channels and data sources still have slower feedback loops compared to others, which create a level of complexity requiring the use of multiple models. The necessary cycle time can also complicate the outsourcing mix as data transfers or legal agreements limit timely analysis. The key to optimizing marketing spend is balancing the application of what is learned from models with the right latitude to encourage innovation and long-term branding.
Much of patients’ data can’t be shared freely due to HIPAA compliance laws, which create a unique data usage problem for this industry. However, patients could be willing to share non-regulated information to the degree that they see value in return. We see this in the form of web and mobile interactions with fitness devices or general wellness information that helps patients lead a healthier lifestyle and access more cost-effective care.
Patients are each on their own individual journey, which may be an annual cadence related to enrollment, or a more frequent cadence receiving communications around care for a chronic condition. It’s important for marketers to collect internal data on member/patient behavior and known preferences, as well as external types of data. For example, life events such as reaching a certain age, moving, or changing family status could trigger certain campaigns. It should be a priority to create a unified and progressive view of the individual patient, customer, or prospect.
Data Quality Management is Essential
With data coming from so many different sources, it’s imperative that healthcare organizations employ data quality management and compliance best practices to better serve patients while protecting their privacy. If they find they have data gaps, they must quickly identify the missing data and determine how to capture and deploy the data needed. To effectively manage all of their data, organizations should deploy the tools that can ingest all types of data in any cadence and then standardize, household, and create. This allows them to persist for a unique key for each patient, resulting in a holistic and up-to-date view of a patient (a “golden record”). This complete view enables organizations to easily determine the appropriate use of data to help them meet their patients’ needs and accurately track where and how the data is used.
Not all data is relevant—in fact, most of it isn’t. Recognizing the fit between the data sources, the analytic tools, and the problem you want to solve is paramount.
However, it is important to remember the basic tenet of data analysis that necessitates formulating questions as the first step before data sources are chosen and analysis begins. Only with clear objectives and strategy can one determine what to measure—and what to ignore.
Once the problem is identified, then the plan can be mapped out and decisions made on what data sources to include and what analytic methodology to apply.
How Data Can Unlock New Customers
Consider the common use case of targeting prospects in a promotional campaign in which the key success metric is the click-through rate to the destination website. Previous segmentation analysis may provide clues on what characteristics are predictive of the desired online behavior. However, in this case, it may be beneficial to look beyond internal customer data and readily available prospect lists as data inputs.
For example, data science techniques applied to a large-scale database may uncover potential new customer segments that seem counterintuitive at first blush because they don’t align with previous assumptions about the ideal prospect. But, when strong correlations are found, marketers can test whether the new segment, in fact, provides good conversion rates.
Marketing and analytics are similar in that you need both a strategy and ongoing iteration. Applying tools aimlessly to existing data and testing all possible correlations is pointless—that’s not sound strategy. But when tested properly on carefully chosen data sets, your tools could mine data for unexpected patterns that inform and enhance your marketing programs.
Getting to the most important and relevant information to calibrate marketing programs and improve commercial success is one of today’s top challenges. The good news is that next generation technology solutions give us sophisticated ways to segment and target based upon key characteristics such as recency, relevance, volume, and more. Marketers should look for solutions that can determine which customers interact with which marketing materials and through which channels so that you can improve your messages and your targeting mechanisms. These capabilities depend upon having the right data sources—both internal and external—and having those fully integrated into a single view.
Choosing the Right Data Platform
It goes without saying that your data analytics platform needs to give your business users the information they need in a timely fashion, but you also need to quickly generate insights. You can’t integrate three different platforms or wait for someone else to ask the business questions. What you need in a complete analytics platform is: 1) Integration of all of the essential data; 2) Regular or real-time updates of all data sources; 3) Pre-built business questions tailored to your role; 4) A simple user interface that gives you the insights you need at the right time.
Data’s Role in Customer Engagement
Everyone in healthcare has a single goal—getting the right treatment to the right patient. For biopharma companies this means understanding who the right physicians and/or institutions are, and how they are interacting within the treatment and/or disease area. Treatment trends, influence within the medical community, and leadership positions among peers are just a few examples of data elements that can increase our understanding of a customer, and help us target them better with the right channels—personal or non-personal—and messages.
We have all witnessed the revolutionary impact data has on all aspects of everyday life. Big data and data science continue to capture the imaginations of business. However, many marketers struggle to make sense of the enormous amounts of data available today.
Terminology is part of the challenge. For example, there is no consensus on which data are actually “big.” Additionally, a lexicon of exotic-sounding data science terms (artificial intelligence, machine learning, neural networks) are frequently referenced but not consistently well understood. Even common terms and algorithms, shared by different fields of computer science and mathematics, can lead to misunderstandings when applied very differently.
So it’s understandable how all this can seem overwhelming. For marketers trying to convert data to value, consider these simple truths:
- While it’s important to collect various types of data, marketers should be careful not to overlook all of the data they already have right in front of them.
- Focus on obtaining the most complete customer profile possible—a “unified customer view” across touch points can help immeasurably.
- Data only deliver if they help drive smarter actions—a well-designed dashboard will not only simplify meaningful metrics, it will also provide context and interactivity to improve actual decision-making.
- Amazing, open-source data analysis and visualization tools are available today—many even outperform proprietary vendors.
- Turning large volumes of diverse, fast-moving “big data” into meaningful insights will likely require different approaches than those designed for conventional structured data.
Remember, analyzing data does not have “one right approach.” But just relying on the approaches and tools you already know can be both good and bad. Find a partner who understands the convergence of marketing, data, and technology for more innovative thinking.
Marketers looking to utilize insights from big data to guide their brand strategies must consider both the data source and the quality of the analytic methodologies being employed. While healthcare-related analytics are diverse, they are also highly specialized. So, in order to derive maximal benefit and extract actionable insights from large healthcare-related data sets, firms specializing in healthcare analytics must have expertise in both data science and healthcare.
Understanding the patient journey is fundamental to biopharmaceutical marketing. Machine learning-based algorithms can harness disease-specific patient encounter data to power analytics that provide marketers with important insights that would otherwise be missed and that can be leveraged to gain a commercial advantage. Similarly, clinical network mapping can provide a better understanding of patterns of patient flow in complex disease states.
Analytics Helps Achieve More With Less
Powerful analytics such as clinical network mapping can elucidate patterns of influence of certain highly influential clinicians on the behavior of other providers, thereby having a ripple effect on the care of many patients—far beyond their own practice. A sophisticated marketer of a brand with a compelling clinical value proposition can harness the power of such analytics. For example, in their promotional, peer-to-peer communications initiatives, this method can be used to help improve patient care and efficiently drive brand utilization through an entire clinical network.
In today’s “do more with less” environment where commercial success must be attained with fewer resources than ever before, marketers can leverage the right healthcare analytics to achieve their business objectives.
Analytical-driven insight from mining big data will become the basis for increasing operational effectiveness, driving productivity gains, and new developments in marketing innovation that will lead to margin improvements.
But how do organizations begin their journey into this new world? From a pragmatic perspective it’s about experimentation—learning new skills and exploring boundaries of how to overcome the challenges of extracting rich commercial insight from fusing the wide variety of meaningful datasets and applying the right types of analytics that will lead to these new commercial insights.
The True Potential of Data
Think of a world where we have curated a dataset that has fused together sales, CRM, claims digital channel interactions along with socioeconomic and demographic data. Predictive models will mine the data to identify which segments are most likely to respond to your promotional efforts and identify factors to improve adherence. Regression models will indicate what the right level of sales calls, sampling, marketing spend, and channel allocation to generate the optimized returns. Demand forecasting models will indicate what the optimized inventory levels should be. And global market access data and optimization models will navigate through the complexities of global pricing rules to reduce global price erosion caused by reference pricing.
Those are just a handful of examples that could result in tangible and demonstrable returns. The curated data set in the future will exponentially increase with the inclusion of social media insight and healthcare sensor data driving a new era of personalized insight.
These new dynamics will introduce the need to overhaul and fully modernize the IT analytical platform to make it fit for purpose. Big data and big analytics will unite the CMO and CIO in their common quest to search for operational efficiencies and margin gains.
Marketers have more data at their fingertips than ever before, thanks to the different tools that they have available to touch base with their various audiences. One data source that’s growing exponentially in importance is social media, as more users prefer to interact with healthcare organizations via Twitter and Facebook rather than phone or email. These interactions can give your organization invaluable data about public perception, customer satisfaction, or even general trends about patient habits.
Data Platforms Right for Marketers
Marketers are not data analysts, so it’s important that whatever analytics platform they choose is a combination of simple, intuitive, yet fully powered. You don’t want to give marketers—who are the first point of contact with both potential and current customers—an inferior solution to analyze all of their available data.
Another important factor to consider is the ability to connect directly to all of your data, no matter where it’s located. Many analytics platforms don’t offer the ability to connect to cloud-based and on-premises data, which could severely handcuff your marketing team’s ability to make data-driven decisions. You want to be sure that you can evaluate your social media data right alongside your email marketing or website visit data without a lot of coding from the IT team.
Improving Customer Relationships
Big data, and the analysis of that data, can give marketers valuable insights into customer relationships, identifying which customers may need a more personal approach than others, and even predict what an interested party may need before they request it. These types of approaches can build a deeper sense of trust in your organization from the customer perspective, which generally leads to increases across all facets of the business.