Data governance

Data Governance:
Towards an agile approach

These last few months, it has become more and more difficult to attend a meeting without hearing the expression data governance. However, this subject is nothing new! Be that as it may, with the arrival of Big Data technologies: data and its use become the cornerstone of approaches to innovation. An old subject evolving in a very new context…

Data and governance: One can’t be without the other

The data craze over the last years is such that enterprises invest a lot of time and money to try to break down data silos and to unify their asset thanks to new, ever more efficient, and less costly storage infrastructures.

Nevertheless, enterprises understood rather quickly that the promise – to innovate through data – was going to be much more complicated than previously expected. Despite the latest technological advancements, data are still scattered on both sides in the enterprise with a militant legacy. New storage systems implemented are, ultimately, “only” additional technical stacks in the enterprise’s IS landscape and don’t allow, on their own, to manage data’s life cycle, guarantee rules allowing the best data usage and thus, maximize the creation of data value. We are talking about data governance here.

The objectives of data governance

In the pursuit of innovation, enterprises are rethinking their organizations to move towards a “data-driven” culture. Information systems must become the profession’s strong arm by placing refined, secured and quality data at the center of strategic decisions.

To achieve this transformation, organizations construct what we call data governance. This project pursues quite clear objectives, among others:

  • Ensure metadata management (technical, operational, or even business) and data documentation.
  • Simplify data access and facilitate their use by as many collaborators as possible.
  • Ensure data quality and integrity.
  • Manage data security: Supervise data collection and their use, especially when it comes to personal data.

A strategy of agile data governance

The way to approach the subject of data governance is evolving. Our experiences have brought us to promote data governance based on the following four pillars:

  • Non-invasive and post hoc: Data governance should not be an obstacle to innovation in your enterprise. Metadata collection and aggregation of an enterprise’s datasets, after their creation or update through various pipelines, allows you not to interfere with the owners of datasets or their users.
  • Automatic and connected: The automation of metadata collection and governance KPIs allows your tools to accurately reflect the reality. On the other hand, this automation guarantees that such governance is up-to-date and ensures upscaling.
  • Bottom-up et collaborative: A strategy of bottom-up data governance wants to put individuals and their interactions in front of processes and tools. An approach to data governance cannot be successful, which involves all the collaborators in an organization, thus benefiting from collective intelligence.
  • Iterative: Construct data governance in stages to correspond as close as possible to the company’s expectations and to its operations. The adaptation to change must be at the heart of an enterprise’s data governance strategy.

Such an approach can be successful where many larger “data governance” initiatives have failed.

bottom up and top down approach

Agile data governance conclusion

The same as how software developments have gradually shifted away from traditional methods (V-model, Waterfall, etc.) to agile methods, data governance must be rethought.

Such an approach is not only iterative but also applied incrementally to your data governance strategy allowing greater flexibility, necessary to take into consideration the ever-increasing complexity of your IS.

white paper : Why start an agile data governance?

The question is not about moving towards becoming a data-driven organization, but how.  We recommend the implementation of an agile, collaborative and pragmatic data governance.

Download

Leave a Reply

Your email address will not be published. Required fields are marked *