Virtually all but the smallest organizations are cognizant of the need for Data Governance and are at various stages of assembling either formal or informal governance measures. How well they do so in 2014, however, largely relates to their ability to cope with the most eminent trends affecting this area of Data Management: Big Data, mobile technologies, Cloud computing, and Master Data Management (MDM).
Big Data Governance
Two of the key issues at the heart of an organization’s ability to embark on a Big Data initiative and to do so successfully pertain to both analytics and governance of such data, which could be as simple as social media sentiment or as complicated as incorporating ceaseless sensor data. Due to the rapidity at which insight from Big Data is able to be derived, the term Big Data Governance is almost an oxymoron since it pits what Forrester’s Michele Goetz identifies as two conflicting interests: the rigidity of time-consuming measures used to trust data and the flexibility of the business user to make real-time decisions.
Effective Big Data Governance, then, ultimately serves as a model for effective governance in the coming years in that it enables such trust relatively expediently. The practical differences between Big Data Governance and the governance of other types of data include:
- Data Sources: Organizations should focus more on establishing the reputation and credibility of Big Data sources, especially in terms of sentiment data gauged from the Internet.
- Metadata Management: Semantic technologies can provide viable means of reducing all forms of data to simple descriptions that are comparable, so that common terms and definitions can apply to Big Data as well as that throughout the enterprise. Ensuring commonalities between the terms and definitions of unstructured Big Data and other data is crucial to Big Data Governance.
- Unwanted Data: Governance rules and responsibilities must be extended to both unwanted and wanted data, since it may take time to discover the value of certain types of data analyzed by Data Scientists. Data Quality standards need to be adjusted for the same reason, since the variation of (unstructured) Big Data implies different facets of consideration for what renders it quality.
- Compliance: Big Data from public sources frequently has specific regulations about usage and length of storage that needs to be ascertained, in addition to restrictions of conventional regulatory agencies.
Master Data Management and Big Data
One of the most rapidly growing trends regarding MDM (beyond the growth of multi-domain MDM systems) pertains to its incorporation of Big Data and Big Data sources. Gartner’s Bill O’Kane remarked:
“…there is an emerging demand to link data from social networks like Facebook, Twitter, and LinkedIn to an organization’s customer and product master data; to fulfill business requirements such as enhanced search capabilities, aggregate sentiment analysis and monitoring, and…to link master data to GPS-generated location data to enable context-aware business operations and decisions”.
2014 should see the continuation of MDM product vendors developing software designed specifically for Big Data within conventional offerings features and unveiling products that link to popular Big Data platforms including Hadoop and NoSQL stores. Current examples of such products include the ability to apply various MDM functions to particular data sources, as well as the ability to connect Big Data sources by workflow or data element – both of which provide the high degree of specificity required to readily apply sentiment data to master data.
Such an incorporation of Big Data into master data shifts governance concerns from specifically relating to Big Data Governance to the more established realm of the governance of master data. This approach also prevents the development of a silo culture and reduces data quality issues such as redundancies or incomplete data.
Research firm J. Gold Associates estimated that “within two to three years, 25% to 35% of business users will employ a smartphone exclusively and abandon fixed-line phones.” This prediction alludes to the increasing rate of adoption of mobile technologies, not only for conventional communication purposes, but for enterprise-wide concerns such as analytics and Business Intelligence. There are several implications regarding the latter trend that directly relate to Data Governance.
The principle point of governance for mobile devices exists at the policy level, and is underscored by the fact that few organizations have specific policy related to the mobile trend which is still emerging, and is not yet pervasive across organizations. As such, there is frequently a multitude of different mobile devices employed – with different capabilities and means of accessing and interacting with company data – and a scarcity of policy to regulate usage. This sort of autonomous mobile usage for analytics can compromise both data quality and security, two not entirely unrelated areas of concern.
Expect organizations to substantially refine governance policy of mobile devices. A Gartner report alludes to this trend and states:
“Enterprise policies on employee-owned hardware usage need to be thoroughly reviewed and, where necessary, updated and extended. Most companies only have policies for employees accessing their networks through devices that the enterprise owns and manages. Set policies to define clear expectations around what they can and can’t do. Balance flexibility with confidentiality and privacy requirements.”
Cloud Management Platforms
Some of the governance concerns that exist for mobile and Big Data technologies from public sources apply to Cloud-based data sources, particularly in terms of security. Still, with more and more systems and data sources accessed through the Cloud, the primary governance issue related to this point pertains to data integration, especially considering the different varieties and degrees of specification of clouds available to users. These issues may be exacerbated by bandwidth and latency problems.
A potential solution to the issue of integrating data through the Cloud with that in the physical enterprise are the use of Cloud Management Platforms (CMP), which should get greater attention from vendors for developing governance capabilities related to resource management, workload optimization and cataloging. According to Gigaom Research: “The interest in CMP is driven by the need to leverage many different types of clouds, including public and private clouds, to provide the ultimate cloud platform for the enterprise”.
Ultimately, the future of Data Governance lies in assuring that data is trustworthy, relevant, and timely, while contending with the ever increasing and autonomous methods of accessing and utilizing such data. Virtually all of the trends identified in this document involve structuring and formalizing the accountability of contemporary and future means of accessing data. Without them enterprise-wide results could truly be a free-for-all with all of the headaches associated with poor data quality, inconsistency, and redundancy which have plagued organizations for years.
Big Data’s influence on governance is considerable and largely represents the duality between controlling data and serving business users in real-time. Refinements and innovations in MDM can assist with Big Data integration and place sentiment analysis readily alongside valuable customer and product data. Cloud Management Platforms can help account for the increasing reliance on the Cloud for contemporary enterprises, while governance policies for personal mobile devices can aid in ensuring that this technology abets instead of hinders organizations. Failure to account for these considerations will likely result in either not utilizing the most effective technologies that an organization’s competitors are, or doing so in a way in which there is no advantage gained.