The Need for Effective Data Governance in Clinical Trials: Part 1

Prakriteswar Santikary, PhD | |

Every organization, including ERT, wants to be data-driven and to monetize its data by building revenue-generating, smart data products. Realizing this potential of enterprise data, however, requires more than just building an enterprise data lake, warehouse or operational data store.

Why data governance is needed

While a modern and scalable data platform is absolutely needed to create these data products ─ including making them intelligent via Artificial Intelligence (AI) and Machine Learning (ML) ─ products must be meaningful, consistent, of high quality. Most importantly, it must be governed and cataloged such that users, both internal and external, trust the quality of the data integration and lineage.

Effective data governance is about specifying the decision rights and accountability to encourage desirable behaviors in the use of data. It touches practically every part of the data management process down to individual technologies, databases and data models (Figure 1).


It affects the processes people use to create and retain data, and how one can replicate these rules within applications to help make smarter decisions faster.

Continuous quality management becomes key, hence the need for effective data governance that works in your organizational environment. While there are multiple ways to establish and operationalize data governance, the key is that the operational model must fit your needs and organizational maturity.

Poor data management leaves businesses vulnerable

On the flipside, ungoverned, unmanaged and un-cataloged data leave businesses vulnerable as users don’t know:

  • Where the data lake is getting sourced from
  • How far the data has traveled before landing into the data lake or data products
  • Whether any business rules are applied to it
  • With whom they can share this data
  • If the data lake or the data product is certified, verified and validated for quality and customer acceptance.

Moreover, regulatory, security, privacy and compliance risks get magnified without effective data governance. Inaccuracies in data inside data products result in unhappy customers and even loss of customers.

In other words, an enterprise data lake without effective data governance becomes an enterprise data swamp very fast, losing trust and market viability for monetization.

However, with well thought-out structure, processes, and rules around enterprise data and its lifecycle within each data product across all data platforms, data governance can deliver real value to the business (Figure 2).

Figure 2: Effective Data Governance Delivers Business Value

In conclusion, effective data governance provides a systematic structure and management to data residing in every data product, making them more accessible, meaningful, reliable and trustworthy.



Prakriteswar Santikary, PhD is the Vice President and Global Chief Data Officer at ERT.