What do we mean when we talk about data governance?

In my experience if you ask a dozen data management professionals you’re likely to get a dozen different answers. Ask those who don’t deal with data and you’ll probably get a dismissive shrug or a mild gripe about bureaucracy and form-filling.

Simply put, it’s neither a well understood nor an engaging subject, yet the reality is that most of us have felt the fallout of poor data governance in one way or another. From local copies of data hidden away in spreadsheets to the perennial disagreement over what a course is. HESA is no different and has wrestled with data governance, locally and on a sector level over a prolonged period.

At the most basic level, successful data governance means managing data to ensure that you get the most out of it. Or, to put it another way, that it meets the needs of its users in the most efficient way. This concept should feel familiar to anyone who has a responsibility for managing other corporate assets. Data is, after all, a type of asset (what type it most closely resembles is something of a hot topic); and there is a lot we can learn from other asset management disciplines. The concept of efficiency is especially important in HE, given the ‘cost’ falls largely on the providers, whilst the ‘benefit’ is mainly felt by the customers. Whilst good governance must be always be underpinned by the anchors of common standards and transparency, what is very clear to me is that governance is first and foremost a people and culture proposition; you could say that governance is a state of mind.

Governance is first and foremost a people and culture proposition

To thrive in this ever-changing world we find ourselves in, we at HESA are currently working towards two broad objectives for data governance: firstly, to design a transparent and sustainable model for managing data collections in the new world of Data Futures, that acknowledges and deals with the fairness of burden. Secondly, to implement the Higher Education Data and Information Improvement Programme’s (HEDIIP) blueprint for governing the redesigned HE Data Landscape that also helps the sector to move forward with us. Data Futures represents - for me - a ‘game changer’ in terms of the data governance capabilities we need to have.

When we factor in both the breadth and subjectivity of data governance (as those who have had the good luck to peruse the DAMA International Guide to Data Management Body of Knowledge, or DAMA DMBOK to those ‘in the know’) to the complex HE environment we operate in, the size of the challenge becomes clear.

For the Data Futures Collection Governance project within HESA, it means that we are principally looking to define robust and repeatable processes, supported by systems that provide reasoned assessment of the cost of change (this is no small task). This improved understanding of the level of burden of data collection enables the decisions that are best for the sector to be taken in a co-owned and collaborative way by those affected. To put this into a more familiar context, the scope of data governance I am talking about covers the activities around record reviews and specification changes, quality assurance, and the curation of the information contained within the coding manual (including coding frames and derived fields). All of this comes with the new dimension of controlling the data models and data language artefacts that were created under HEDIIP, representing the sector model of best practice.

To support this, we need to move towards an approach that allows informed and intelligent conversations; a core part of which must be on articulating the ‘justified burden’ of data supply. We need to have these substantive conversations alongside the unfortunately necessary - but proportionate - ‘bureaucratic form-filling’. This allows the system to operate efficiently, providing the lowest cost of supply while supporting the needs of the customer.

Visit the Collection Governance project pages - we are seeking feedback and further queries via survey by Friday 18 August.