By Ian Ferré, Associate Director, Infoverity, Inc.
Often, companies think of data governance in a parental way, something that’s a nag and a hindrance to freedom and speed, and a burden on their teams. They may even be inclined to dismiss it. But while data governance naturally involves rules, audits, policies and other guardrails, the larger mission of data governance is not restrictive at all. Instead, it can be empowering and can ultimately give a business more room to experiment and make swift decisions while being confident in the underlying data.
When thought of as a mission to create trusted data to make a business run better, data governance becomes exciting. By understanding what data is needed, making sure it is collected, improving quality starting at the source, providing ways to create and publish purpose built collections of data, data governance creates data that the business can actually believe in.
The alternative path in which data governance is a second class citizen is not pretty. Too frequently, companies or departments end up engaged in “data brawls,” in which people are tempted to attack the data if something goes wrong or if an analysis reveals an unwanted truth. Such data brawls take the company’s focus off of root-cause issues surrounding the data and how the business is being run.
This blog examines the essential role of data governance, and why it can be a strategic enabler, one that improves the quality and reliability of enterprise data and, rather than allowing data to be a chronic source of doubt or confusion, can actually restore and foster faith in the data being used to make business decisions.
Focus on the Business Outcome of a Data Governance Framework, Not The Limitations
To get data right, companies should start with the particular business outcome they want to achieve. Whether that is upselling or cross-selling a specific product to customers, improving customer retention, or some other goal, the business outcome is the cornerstone of the data governance process. Once the business knows what needs to be accomplished, it becomes easier to examine what’s working — and what’s not, both in the data, and the business as a whole. From there, companies can put processes in place that support the desired outcome.
While it may seem daunting and you may not know where to begin, there are concrete steps you can take to make big improvements in data governance.
Data Governance Best Practices: Four Ways to Make the Process Work Better for Your Company
Examine your data pipeline. From there, companies should focus on their data flow, understanding where data elements are created, who created them, why they were created, and what governance standards may be in place already. Ensuring the quality of the data at its source prevents problems from arising further down the pipeline as data is cleansed, harmonized, extracted, analyzed and delivered to business users.
Prioritize. What are the most important questions you’re trying to answer? What are the most pressing decisions you need to make? What data is associated with them? Look at what you need or what isn’t working, the level of effort it would take to fix it, and the impact better governance would have. You can then create a roadmap that shows you where to focus your attention and resources first. By starting on smaller projects with big impact, you can build confidence and start to create a culture that embraces data governance.
Create standard definitions and document them. Definitions are critical to making any meaningful improvements. For example, if you do business in Puerto Rico, is Puerto Rico classified as a state or a territory within your data? Are you following international address standards or U.S. address standards? As business users know well, often such obvious problems exist in data sets. A lot of poor data can be fixed by taking the time and effort to define and enforce sound business rules.
Take an incremental approach rather than trying to boil the ocean. Once you prioritize your worst hot spots or biggest opportunities, break the problem down to an atomic level. Problems can be resolved in many ways — better processes, system controls, or retroactive data cleansing and stewardship. For each opportunity, figure out what the short-term and long-term fixes are and implement your plan.
Pathways to Better Data Quality
A large part of data governance is about improving data quality in three ways. One way is to improve the quality of the data at the point of collection. Another way is to cleanse and transform data within integrations between systems (or better yet, perform transformations in the source system). A third way would be to understand the sources of variation in the data and to solve these at a process level by asserting standards that minimize or remove the possibility for variance. Data governance isn’t an exact science and there are no perfect answers, and many businesses end up using some combination of the three.
The human factor in data governance. In addition to technology and business rules, there is a human element to data governance that can’t be overlooked. People must be trained on how and what data to enter. Successful governance means helping people understand why the data is being collected, how it’s being used, why it has to be accurate, and why it can’t be duplicated. It also means making it easy for people to follow the agreed upon rules. They need help understanding implications for the business when data governance fails and when it’s working well. And they need to understand where they fit in the broader process, so that they can truly understand why data governance matters. There’s a reason why the classic model of “people-process-technology” continues to stand the test of time. Companies can’t function fully with just one or two, they need all three working in harmony. The same holds true for data governance.
When business outcomes are clear, human involvement causes fewer problems because people are not just doing governance for the sake of data governance. Governance for the sake of governance doesn’t work because it makes it harder to access data than it should be. Data governance, when done correctly, is in the business of improving access to information, not taking it away. The goal of data governance should be to allow those in a business to have access to the right information at the right time that’s going to help them drive business decisions and ultimately remove errors.
Data Governance Policies: How to Better Measure Results
To ensure data governance programs are having the impact on the business outcomes the company desires, metrics should be put in place to assess progress. This could include measuring what can be done now that couldn’t have been done with bad data, such as new recommendations and cross or up-sells for customers. Or examining how offers can be better tailored due to better data, driving an increase in revenue.
But companies should also have data quality metrics that measure factors like the consistency, completeness, accuracy, validity, and timeliness of their data. Data should be consistent, so that if data is in more than one place it doesn’t vary. Data should also be accurate and valid so that it matches the source data. Metrics should also be in place to ensure data sets are complete, and that data can be generated, accessed, and used in a timely fashion.
Ultimately, data governance sometimes gets a bad rep because its true mission is obscured. Without the proper context, few people are excited about establishing rules and policies to control data. But these rules and policies are not the point. The trusted data assets that are created by good data governance and the mistakes and costs that are avoided are the real victory.
Founded in 2011, Infoverity is a leading systems integrator and global professional services firm driven to simplify and maximize the value of their clients’ information. Infoverity provides MDM and PIM Strategy and Implementation, Data Governance and Analytics, Content Management, Data Integration, Enterprise Hosting, and Managed Services that help large enterprises in the retail, consumer goods, manufacturing, financial and healthcare sectors. Infoverity, a 100% employee-owned company, is on the Inc. 5000, recognized by IDG’s Computerworld as one of the Best Places to Work in IT, as a Wonderful Workplace for Young Professionals and as a “Best Place to Work” by Inc. Magazine and Business First. Infoverity’s global headquarters is in Dublin, Ohio, the EMEA headquarters and Global Development Center is in Valencia, Spain. Additional offices are located in Germany and India.
For more information on Infoverity solutions, contact us today.