More and more organizations are seeking guidance to help them understand how to harness the power of big data. These organizations want to know how to efficiently manage massive volumes of complex (structured, semi-structured, and unstructured) data and how to leverage analytics (predictive, social, mobile) to gain better insights to improve operational efficiency, increase profits, and gain a better competitive position in the marketplace. Organizations across nearly every sector are looking for ways to tap into data that was previously trapped in unstructured sources, such as text documents, email, and social media sites, and leverage this previously untapped data to turn poor business decisions made using haphazard guesswork into well-considered and successful business decisions that improve overall performance.
Yet, while forward-looking insight is the ultimate goal, organizations must understand the data they collect and store. This data has to be identified, acquired, organized, filtered, and cleansed, then integrated and stored before it offers real value to the end business consumer. This basic “blocking and tackling,” which I have preached for years as a business intelligence professional, is a critical underpinning for any future analytics initiative. This begins with a clearly defined and agreed-upon data governance strategy.
What is Data Governance?
Data governance is the process of creating and agreeing to standards and requirements for the collection, identification, storage, and use of data. This should not be viewed as optional with any data-driven project. A data-governance initiative helps the organization set rules, policies, standards, and procedures, and define roles and responsibilities with respect to the overall management of data. Effective data governance allows for the efficient integration of new data sources and helps the organization realize value from this data – such as the ability to aggressively pursue new market opportunities and identify and capitalize on emerging business opportunities. Other key benefits of an effective data governance program include – but are not limited to – reduced cost for data storage, reduced cost for rework often associated with poor data quality, increased confidence in data quality (often a result of improved data consistency), improved performance of technology solutions, and enhanced data security.
Data governance has not changed dramatically over the years, but the types and volume of data being collected, stored, and used for analytics has changed significantly. Big data governance requires governance over many different types of data (including metadata, or data about the data), not just what’s in the legacy systems or relational databases. This requires a new understanding of the methods, processes, and tools that must to be deployed to deal with this big data. Organizations spend a great deal of time, money, and human capital developing big data programs that involve implementing data management solutions and ironing out the integration processes needed to tie it all together.
It is estimated that 80 percent of the data being created today is unstructured. This only exacerbates the issues as organizations begin to tackle their big data challenges around weaving together structured, semi-structured, and unstructured information. While data can provide game-changing insights to run the business, key leaders are left wondering what data they have, what data is needed to answer key organizational challenges, how the data should be integrated, where the data should be stored, what processes should be in place to assure data accuracy and security, and who is allowed to view the data. This is where a robust data governance plan can help and create order from chaos.
A data governance program must be in place at the outset of any large-scale technology project so that the resulting insights can be trusted to help the organization achieve value from the investment being made. This will ensure the organization gets the right people involved, can define and adjust the relevant processes for data management, and deploys the right technology solutions to address the needs of the business and the complexity and volume of data being managed. Not having a data governance program in place can result in misalignment between the data and overall business strategy and reduce trust between business and IT due to a fundamental lack of trust in the information being provided.
One of the most important aspects of data governance is alignment between IT and business. Having clearly defined roles and responsibilities and objectives understood from the outset is paramount to the success of any governance initiative. Agreement on the use case for this type of initiative and business value that is being sought will help gain support and funding for the initiative within the organization.
As for initial tactical steps that an organization should take, that is largely dependent on the organization’s data maturity, ability to manage data, and understanding of the objective(s) of the key consumers of the data. Most data-governance programs begin with agreement between IT and business on the business needs/goals/expectations, the approach to data management, the scope of the work, the business and technology requirements, the design of the technology landscape, and the deployment of assets to realize/implement the solution.
Data-driven organizations of all sizes, across all industry sectors, are seeking better ways to manage the growing volume of data and create and disseminate relevant, accurate, trusted, and timely information. In this era of big data, an effective data governance program makes this possible. When governance is layered over the framework of a data analytics platform, the result is a holistic understanding that improves the company’s ability to manage and measure ROI. The more that companies invest in strategic analytics, the more they will need a robust data-governance plan to extract results from their data and ensure immediate and long-term success.
Scott Schlesinger is a Principal within Ernst & Young LLP’s National IT Advisory Practice, and serves as EY’s Americas Leader for Business Intelligence and Information Management.
The views expressed herein are those of the author and do not necessarily reflect the views of Ernst & Young LLP.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.