Manage Learn to apply best practices and optimize your operations.

Common data vocabulary helps companies globalize governance, BI

Strong data governance is needed to maximize the value of business intelligence efforts -- and that starts with a common set of data definitions, consultant Wayne Eckerson says.

I know we're in the age of big data, but the key challenge facing most of the large companies I work with is a lack of a common data vocabulary. Internally, such organizations are a proverbial Tower of Babel, communicating at cross-purposes with data and performance metrics that are inconsistently or ambiguously defined. Departments and business units don't speak the same language, which undermines operational efficiency and competitiveness. What's missing is a strong, vigorous data governance program for managing data as a critical business asset.

My clients are data-driven companies with substantial investments in data warehousing, business intelligence and master data management. But forget about big data for the moment: Many of these companies struggle just to manage what I call their small data -- information from enterprise resource planning, customer relationship management and other run-the-business systems.

The problem is that they're fractured organizationally, with each business unit or regional operation managing its own applications, systems and customer databases. Although that empowers business units to tailor the delivery of products and services to customers, it wreaks havoc on global data and, subsequently, a company's ability to coordinate resources efficiently. In this state of affairs, companies are optimized locally and suboptimized globally. That eventually erodes their ability to compete effectively in the broader market.

Instead, organizations should strive to optimize both local and global processes. Doing so involves establishing global data standards that define the meaning of shared data and processes for managing them. That requires representatives from all the business units to spend time working with their counterparts to hammer out a common vocabulary and data governance framework.

Defining data a worthy effort

But if done properly, data standards neither constrain business units nor hamper their ability to serve customers. On the contrary, well-governed standards foster greater collaboration among business units. That makes it easier for customers to do business with the company as a whole, which benefits each business unit in the long run.

Of course, creating a common data vocabulary is easier said than done. Two common approaches to implementing data governance programs are the executive fiat and the Trojan horse. One is a top-down mandate driven by the CEO or another top executive; the other a bottom-up deployment tucked inside a strategic IT initiative. Neither is ideal on its own, though -- let's look at why.

Often in large companies without unified data standards, the CEO eventually becomes frustrated by the lack of coherent and consistent data to answer even basic questions -- such as "How much did we do in sales yesterday?" or "How many customers do we have?" or "Why is there an above-average return rate on our products?" The CEO -- or another corporate exec with enough clout -- can initiate and fund a governance program to create and manage a data vocabulary and enforce its use in all operational systems and BI, reporting and analytics applications. Ideally, the program is designed to get input and gain buy-in from every part of the organization.

But frustrated executives often take shortcuts. They expect a mandate to be sufficient to change how data assets are managed throughout their companies. So they don't put in place a change management process to give business unit managers a chance to provide input on the shape of the program and the data standards themselves. And without sufficient support from mid-level managers, a data governance program is unlikely to ever get off the ground.

Stealthy advances on data governance

The Trojan horse approach embeds data governance procedures into a technology project, such as the deployment of a new CRM, compliance reporting or risk management system. By piggybacking on a strategic project, governance proponents can build support for precise data definitions and standardized data management processes across business units -- and eventually win executive support and funding.

For instance, one of my clients is a global medical equipment manufacturer whose CEO wants a performance scorecard to better manage the company, which consists of dozens of autonomous business units, some of which are independent legal entities. There was no data governance mandate. But in creating the scorecard, the design team developed an informal governance program to get agreement on the definition of the scorecard metrics and their underlying business terms. The team had to gather senior-level decision makers from across the company in each functional area --  sales, finance, human resources, others -- to hash out definitions, filters, access rights and data quality standards. And now they have to oversee a process to monitor data quality and manage changes to the metrics.

The team now recognizes that the real purpose of the scorecard, and its primary benefit, isn't to provide charts and data to the CEO -- it's to create a common vocabulary that the company can use to do business globally. And it desperately needs that vocabulary to compete against larger, more integrated competitors that are entering the marketplace.

But to improve their chances of success, companies should blend the two approaches. The Trojan horse one often doesn't work without some sort of executive order to clean up data and make it consistent. And an executive fiat should be tied to other initiatives and implemented by focusing on data elements that everyone recognizes are critical to the operation of the company. (Or to their bonuses!)

Don't just let things stay the way they are, though. The reality is that data governance, not big data, is the defining data management challenge of our time. It's one thing to be a data-driven organization; it's another to have a common vocabulary of shared terminology and metrics. But it's a must for competing in today's information economy.

Watch for the next article in this series on the two worlds of data governance.

About the author:
Wayne Eckerson is principal consultant at Eckerson Group, a consulting firm that helps business leaders use data and technology to drive better insights and actions. His team provides information and advice on business intelligence, analytics, performance management, data governance, data warehousing and big data. Email him at wayne@eckerson.com.

Email us at editor@searchbusinessanalytics.com and follow us on Twitter: @BizAnalyticsTT.

Next Steps

More from Wayne Eckerson: BI program directors need a strong mix of business and IT skills

See why he says you should call in a BI SWAT team to rescue failing projects

Get his advice on how to banish inconsistent analytics data from your organization

This was last published in January 2015

Dig Deeper on Business intelligence best practices

PRO+

Content

Find more PRO+ content and other member only offers, here.

Join the conversation

4 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

What's your top tip on how to successfully create and maintain a common set of data definitions in an organization?
Cancel
The method I use with my customers is good old ER Data Modeling. Organize fast data modeling workshops with business users. Build first data models for each functional area ("silo"). Build then a high level data model which covers all main entities across applications and organization units. Data modeling is all about communication. So discuss common definitions, example data and synonyms for each entity and data item. Make extra effort to draw clear, understandable logical data models (no messy database diagrams!)  If organized well, this process can be carried out in a surprisingly short time. 
Cancel
When rolling out your new data definitions, use multiple communications channels to get your message across. Make sure you target your communications to your various audiences. For example, those inputting data might need face to face learning, whilst those using reports may only need to know that the changes have occurred (with a link to the details).
Cancel
Remember: one datum, two data.
Cancel

-ADS BY GOOGLE

SearchDataManagement

SearchAWS

SearchContentManagement

SearchCRM

SearchOracle

SearchSAP

SearchSQLServer

SearchSalesforce

Close