|An Introduction to Big Data Governance|
|Database - Business Intelligence|
|Written by Sunil Soares|
|Monday, 12 November 2012 00:00|
We are drowning in data today. This data comes from social media, telephone GPS signals, utility smart meters, RFID tags, digital pictures, and online videos, among other sources. IDC estimates that the amount of information in the digital universe exceeded 1.8 zettabytes (1.8 trillion gigabytes) in 2011 and is doubling every two years.1 Much of this data can be characterized as big data. Big data is generally referred to in the context of the "three Vs"—volume, velocity, and variety. We add another "V" for value. Let's consider each of these terms:
Organizations must govern all this big data, which brings us to the subject of this book. We define big data governance as follows:
Big data governance is part of a broader information governance program that formulates policy relating to the optimization, privacy, and monetization of big data by aligning the objectives of multiple functions.
Let's decompose this definition into its main parts:
Big data is part of a broader information governance program.
Information governance organizations should incorporate big data into their existing information governance frameworks by doing the following:
Big data governance is about policy formulation.
Policy includes the written or unwritten declarations of how people should behave in a given situation. For example, a big data governance policy might state that an organization will not integrate a customer's Facebook profile into his or her master data record without that customer's informed consent.
Big data must be optimized.
Privacy of big data is important.
Organizations also need to establish the appropriate policies to prevent the misuse of big data. Organizations need to consider the reputational, regulatory, and legal risks involved when handling social media, geolocation, biometric, and other forms of personally identifiable information.
Big data must be monetized.
Monetization is the process of converting an asset such as data into money by selling it to third parties or by using it to develop new services. Traditional accounting rules do not allow companies to treat information as a financial asset on their balance sheets unless purchased from external sources. Despite this conservative accounting treatment, organizations now recognize that they should treat big data as an enterprise asset with financial value. For example, operations departments can use sensor data to increase equipment uptime based on preventive maintenance programs. Call centers can analyze agents' notes to reduce call volumes by understanding why customers call. In addition, retailers can use master data to power Facebook apps that drive customer loyalty.
Big data exposes natural tensions across functions.
Big data governance needs to harmonize competing objectives across multiple functions. For example, the wireless marketing department at a telecommunications carrier might be interested in using geolocation data to drive new revenue streams, such as when a subscriber receives coupons from retailers that are in close proximity. However, the wireline business might be concerned about the reputational hazard associated with reusing subscribers' geolocation data without their consent. Meanwhile, the network management team might want to use this information to address any issues with network performance, such as a large number of dropped calls at a specific wireless tower. Finally, the chief privacy officer might have concerns about the potential for regulatory backlash. In this situation, big data governance needs to bring all the parties together to determine whether the potential revenue upside from the new services outweighs the associated reputational and regulatory risks. The usage of geolocation data for internal network analytics is probably okay, but the other business uses might not be.
Case Study 1.1 reviews the unfortunate events surrounding the Mars Climate Orbiter. We would not consider this volume of data to be "big" by today's standards. However, NASA likely produced the navigation commands by crunching some very big numbers with complex mathematics. If commercial organizations do similar crunching of big data to score a risk, fraud, or propensity to buy, they might incorrectly reject credit card applications or miss customer churn events because scores are misunderstood or applied incorrectly.
Case Study 1.1: Big data governance and the Mars Climate Orbiter2,3,4
In a typical information governance project, the team identifies a business problem, develops a business case, obtains an executive sponsor, defines a technical architecture, and proceeds with the rest of the initiative. However, big data projects are different because of the following characteristics:
As of the publication of this book, governance has taken a backseat to the analytics and technologies associated with big data. However, as big data projects become mainstream, we anticipate that privacy, stewardship, data quality, metadata, and information lifecycle management will coalesce into an emerging imperative for big data governance.
1. The 2011 Digital Universe Study: Extracting Value from Chaos (IDC, 2011).
3. "Mars Climate Orbiter Fact Sheet." http://mars.jpl.nasa.gov/msp98/orbiter/fact.html.
4. "Mars Climate Orbiter Mishap Investigation Board Phase I Report." November 1999.
|Last Updated on Monday, 12 November 2012 00:00|