In life sciences sales and marketing, we are leveraging big data and analytics to drive next best action engagement with healthcare providers (HCPs), to optimize the delivery of patient care and increase outcomes. However, while claims, EHR, prescribing, and quantitative market research, is great for apprising what, when, how, and where, the “why” is still left on the table.
Qualitative information helps round the edges of a strict analytical view of patient and HCP behavior, filling in the gaps as to the motivations and social context related to behavior. Clifford Geertz, an anthropologist, described this type of data as being “thick”. Due to the unstructured nature of qualitative, thick, data it has been historically difficult to synthesize multiple sets and analyze large volumes.
With recent developments in natural language processing, machine learning, artificial intelligence, and neural networks, it is now possible to liberate the multi-dimensional nature of thick data, helping to understand sentiment and context in a new way. All models depend on training data, in this instance educating an artificial intelligence on how people speak about their healthcare experiences derived from observational information. Since human experience is so vast and varied, it takes significant time and resources to build models which are accurate.
Therefore, the limiting factor to successfully elucidating insight from thick data is having access to advanced machine learning models which are sufficiently both deep and broad in their comprehension of the intangible aspects of the healthcare journey, from the perspectives of patients, providers, caregivers, and other stakeholders.
