Data governance that delivers benefits to the business needs to pull together people, process and technology by embracing data ownership, managing metadata, ensuring data quality, and continually improving automation.
These are just some of the guidelines for successful data governance set out by a panel of experts at A-Team Group’s recent Data Management Summit in New York. The panel was moderated by Andrew Delaney, president and chief content officer at A-Team Group, and joined by Ellen Gentile, data quality leader at Edward Jones; Randy Gordon, head of data governance at Cross River; Mark Shainman, senior director of data governance products at Securiti, John Carroll, head of customer success at Datactics; and Derek Wilson, executive director, Lumada Solutions Experience at Hitachi Vantara.
Data ownership
The panel opened with a look at how to embed data ownership across an organisation, which has been something of a problem in the past. Things are changing, however, with the panel noting that business people increasingly understand the benefits of data ownership and want to be accountable for the data they use so that they can take an offensive approach to data management and maximise data assets. As one speaker commented: “Data owners make business decisions.”
Despite this advance, data ownership raises challenges, such as who is creating and defining data, a problem that can be eased by adding criteria to data definitions that can be used across the organisation. The issue of data owned by one person being better or worse than data owned by another is also problematic.
It is here that people, process and technology come in, with upstream data owners making data definitions and, as the data moves downstream, data stewards and data owners taking responsibility for the data and considering its quality. Collaboration, data transparency and workflows help the understanding of who owns particular data and how to work with other data owners.
Turning to the need for data quality, a panel member noted: “Data quality management is at the heart of avoiding data becoming a risk,”. To achieve quality, another noted the need for data stewards to identify critical data elements and create data quality rules for these elements, and avoid boiling the ocean.
Use cases
Considering the use cases of best practice data governance, the panel considered the implications of adding unstructured ESG data. The consensus was that ESG data should be treated in the same way as financial and economic data, and to treat it separately would lose all the benefits of scale previously built into data management and governance.
Data privacy was also highlighted, with one speaker noting, “Organisations must understand what sensitive data they hold and how many systems it is in.” Doing this manually is not scalable. Instead, technologies such as AI and ML can be used to find out where the data is, and controls can be imposed on the data depending on its sensitivity. “Data governance is important here, and metadata needs to be added to govern the data appropriately.”
Technology
Technology plays a critical role in managing huge volumes of data as it scales and becomes more complex, while the creation of metadata allows AI and ML to identify hidden data relationships that are not otherwise apparent. This process is a driver of automation, and as one speaker put it, “Automation helps data discovery and access to the data.”
Closing the technology discussion, another a speaker said: “Technology helps data governance and data governance helps the business. By prioritising metadata, data governance drives value out of data that is not achievable using anything else.”
Subscribe to our newsletter
