In this way, governance is planned and implemented to create a competitive advantage, addressing policy compliance, security, accessibility, and usability in an easy and comprehensive way. This in turn will increase data availability and increase its usability for distributed team members — while maintaining centralized control over risks. While common data governance practices present obstacles for businesses, a combination of these models has the potential to overcome them.
Both data governance models pose challenges
Companies are struggling to manage data at scale and in the cloud. Almost three-quarters of people make decisions in one Forrester Research’s recent poll say they don’t manage most of their organization’s data in the cloud yet. About 80% say they have difficulty managing data at scale. A whopping 82% say forecasting and cost control is a challenge in their data ecosystem, and 82% say confusing data governance policies is a difficulty.
Meanwhile, the volume of data that companies have to manage is mushrooming, and more and more users are demanding more access. “You now have more data coming from many other sources stored in different places,” said Patrick Barch, senior director of product management at Capital One Software.
Organizations want to make this data accessible to more business teams, enabling new insights and more business value. However, many struggle to balance the need for centralized governance of data in the cloud — ensuring comprehensive governance but can clog data access — with a decentralized model that helps industries business control and access to more data and analysis. Decentralization, however, has its own downsides. Different groups may not agree on governance policies. Specific data or data types can get stuck in silos, not all of which are available. Machine learning engineers may lack access to the data they need to build advanced analytics tools.
“Your teams want complete and immediate access to the data and tools of their choice,” says Barch. “You can’t manage everything centrally without becoming a major bottleneck or hiring an army of data engineers, and you can’t completely decentralize management responsibilities without taking on data risks. Is it significant?”
The best of both worlds
However, there is a way to combine centralized and decentralized approaches into a new data governance model through data management federation. Doing so allows businesses to realize the advantages of each, not the downsides.
For example, Capital One adopted this model while the company closed its data centers and moved operations to the public cloud. The company deployed a cloud data warehouse to make data available to business teams widely, but realized that it needed to pay attention to data governance.
“Without good governance controls, you not only run the risk of policy management, but you also run the risk of spending more money than you intended, much faster,” says Barch. “We know that maximizing the value of our data, especially given the amount and variety of that data, will require creating an integrated experience with integrated governance that allows Various stakeholders participate in activities such as data publishing, data usage, data management, and underlying infrastructure management, so that it all works seamlessly together. “