By Tyler Beasley, Executive Director, Business Intelligence
Over the last several years, clinical research organizations have invested heavily in business intelligence infrastructure. Enterprise reporting has replaced manually maintained spreadsheets and datasets. Data warehouses have helped to partially bridge the gap with previously disconnected systems. Dashboards now deliver visibility into nearly every aspect of study execution, from activation and enrollment to fiscal performance and resource allocation.
These sorts of developments underscore the meaningful progress being made in the industry.
From a historical perspective, gathering operational information frequently required significant manual effort. While reports were generated, data definitions were inconsistent and varied between departments, and leaders often made decisions based on incomplete information. Contemporary Business Intelligence programs have helped to address a variety of challenges by improving access to information and increasingly organizational visibility.
Despite this progress, an interesting reality persists: most organizations that describe themselves as “data-driven” still struggle to operationalize data-driven decision-making processes. This does not, however, diminish the value of Business Intelligence; rather, it highlights a critical distinction that is often overlooked within discussions surrounding analytical and organizational maturity.
Availability of information and utilization of information are not synonymous.
Neither the ability to generate data – nor the access to it – necessarily translate into actions or decision making.
The Assumption of Analytical Maturity
There exists within most organizations an implicit assumption that the implementation or application of technology signals successful analytical maturity—this reasoning is sound. With centralized data that is governed appropriately – which is presented through accessible and utilizable reporting infrastructures – decision makers, at least in theory, possess all the information necessary to make data-informed decisions. In response to that, substantial investments are made in data reporting platforms, visualization technologies, and data infrastructure. While these investments are necessary, alone, they are not sufficient.
Business Intelligence initiatives frequently find their focus in solving technical challenges related to availability of data, accessibility, and integration. While these initiatives do improve visibility, they do not address organizational behavior that influences the interpretation and application of information. The purpose of Business Intelligence has never been to produce data, but rather, to improve decision-making. In short, organizations may become exceptionally proficient in producing information while simultaneously remaining less-than-effective at operationalizing it.
This is a fundamental distinction; the purpose of Business Intelligence has never been the production of data or reports but rather, to improve decision-making. In clinical research, integrated site organizations like Velocity Clinical Research demonstrate this principle: centralized data governance, standardized technology deployed across all locations, and unified training and accountability structures create the organizational conditions in which BI data actually drives consistent operational decisions.
The Persistence of Competing Truths
Among the many challenges that Business Intelligence leaders face, one of the most common is the lack of agreement in regard to the meaning of data. Most organizations have experienced some variation of the following: A report is distributed to a leadership team. Questions surface: what system was the data pulled from? How were the calculations determined? Why do the numbers differ? What version is correct?
Admittedly, these questions appear to be technical in nature…at least at first. The reality is that the core of the question is rooted in organizational trust. Data-informed decision-making demands not only access to information but complete confidence in the information that has been presented.
Absent confidence, decision-makers commonly regress to other forms of validation: institutional knowledge or memory, experience, etc., and individual analysis becomes the replacement for analytical evidence. Though data may be accessible in these settings, it rarely functions as the primary catalyst for decisions.
Information Does Not Eliminate Bias
I’ve observed the tendency to assume that improved or increased access to data automatically equates to less subjective decision-making; this is an oversimplification, and the reality is considerably more nuanced.
Individuals rarely engage with data in a vacuum. Business Intelligence often is discussed as though data itself possesses objective meaning independent of the individual reviewing it. However, in practice, decision-makers interpret information through the lens of previous experiences, expectations, and well-established assumptions. Sociologist Pierre Bourdieu discussed this phenomenon through the concept of habitus; that is, the framework through which individuals interpret the world around them. There is a parallel between this sociological phenomenon and Business Intelligence. As a result, one report may yield vastly different conclusions between different stakeholders. While data may be objective, the interpretation rarely is.
This does not represent an inherent failure of Business Intelligence. Rather, it is reflective of the reality that data alone cannot eliminate bias, uncertainty, mistrust, or competing priorities. Organizations commonly discover that the largest hurdles preventing them from becoming data-driven are not related to software or technology, but to culture. Building an environment in which information is trusted, known, and consistently utilized requires commitment rather than technical implementation alone.
Measuring the Wrong Indicators
Business Intelligence initiatives are frequently evaluated based upon easily-quantifiable measures. For example, organizations may track the number of deployed dashboards, adoption rates, data refresh sequences, etc. Though there is no argument that these metrics can provide insight into system utilization, the reality is that they reveal little about overall organizational impact.
This observation calls attention to a larger challenge for Business Intelligence. Organizations tend to prioritize variables that are easily measurable or quantifiable over those that are more difficult to measure. Factors that are most influential to organizational success do not necessarily lend themselves to simple quantification. Trust in data, leadership confidence, institutional knowledge, alignment with stakeholders, and organizational culture pressurize decision making – but to measure or quantify that pressure is a challenge.
That said, this challenge is not unique to Business Intelligence. Researchers have grappled with measuring non-cognitive variables that influence outcomes for decades. Resilience, motivation, belonging, and confidence are widely acknowledged as predictors for success; not only do they remain difficult to define operationally, but they are also difficult to measure with consistency. The fact that they are challenging to measure does not diminish their importance. For example, Velocity’s CARE Councils bring Principal Investigators, operational experts, and site staff together across therapeutic areas to build consensus on protocol design and operational standards before a trial launches. This structure doesn’t eliminate bias, but it creates organizational alignment and a shared frame of reference, which reduces interpretation variance and increases confidence in data-driven decisions.
The same principles exist within organizations. Exclusively focusing on what can be quantified may indirectly obscure variables that are (at least) equally significant but less readily observed. In short, Business Intelligence programs can easily become surprisingly effective at outlining observable organizational conditions while simultaneously remaining ineffective at explaining the underlying factors that produce them. Ultimately, the value of Business Intelligence does not reside in the quantity of reports, but in the quality of decisions enabled.
The Next Phase of Business Intelligence
The clinical research industry has made undeniable progress in advancing the analytical capabilities. Data infrastructure improves, reporting environments mature, and organizations now host access to more information than ever before. These are meaningful achievements; however, the evolution of Business Intelligence will be determined by a different challenge. We should no longer ask whether organizations can access data, but rather, whether organizations can consistently transform that data into meaningful action. To achieve this, technology alone will not be sufficient. There will be requirements for defensible and trusted data governance, defined metrics, organizational alignment, and an overall commitment to embedding analytical insight into everyday decision-making.
In many respects, the future of Business Intelligence will be measured not by the ability to generate information, but by the ability to glean meaning and understanding from it.
