Pulling it all together: a new model, the enterprise data hub, can present a 'single source of truth' without disrupting existing business processes or requiring costly IT reinvestments.

If yours is like most companies today, it has spent millions of dollars purchasing, implementing and maintaining an extensive set of business applications. But if you were to ask a room full of executives, "Those of you who can achieve a comprehensive view of your business, please raise your hands," it's likely you would see few--if any--hands in the air.

Achieving a "single source of truth" is an uphill battle that companies of all sizes face on a daily basis. Executives are making critical business decisions on intuition rather than on credible information, and the financial and other risks of this all-too-common scenario are startling. Further, federal regulations like The Sarbanes-Oxley Act have essentially created a mandate for complete and accurate data.

Whether it's marketing activities, supply chain operations or the finance department, all facets of an organization rely on customer data in some capacity. However, this very data that is recognized as so critical to the enterprise has been recklessly scattered across the organization due to the decentralized organizational structure of most companies. The result is an organization with departmental silos that do not--and often cannot--share customer information. In addition, each department may "cleanse" and standardize the data at each source separately, which may result in inconsistent and redundant processes, along with significant additional costs.

An architectural innovation, the enterprise data hub model can provide a comprehensive, and cost-efficient, single-source-of-truth solution. Using such a model, all legacy system and third-party information is collected centrally in an online repository, presenting one view across the enterprise.

Within the hub environment, the single source of truth can be achieved without disrupting existing business processes or requiring costly information technology (IT) reinvestments. Furthermore, all data quality and data maintenance services can be centrally maintained and managed with the clean and standardized data flowing throughout the organization and available to all users across all departments.

The conceptual breakthrough in the enterprise data hub comes in understanding that certain data entities are shared resources across an organization. Instead of each department maintaining its own version of a customer record separately, all shared attributes describing the customer record are merged into a single master file. This approach allows for all relevant business transactions, activities and interactions to have visibility without the requirement of costly movement of transactional data. The data entities are the cross-reference keys that allow organizations to obtain a single source of truth without having to disrupt the transactional systems that are running the business.

The enterprise data hub is an operational system working in real-time. As data enters the hub, the system automatically begins to verify, cleanse, de-duplicate and merge the information, then synchronize all systems. As a result, all corporate users--regardless of their role or location--use the same accurate, continually updated information.

[ILLUSTRATION OMITTED]

The hub is analogous to a public water system. The water is centrally collected, purified and tested; it is then distributed from the pumping station to all connected users. This is infinitely more logical than circulating untreated water and expecting each user to install a complete water filtration and testing system. With the hub as the data clearinghouse, all source systems use the same standardized processes and "clean" source of information.

Technology-savvy executives may ask how an enterprise data hub is different from a data warehouse, as both approaches sound similar. An enterprise data hub collects and cleans shared business entities across multiple transaction applications, such as customers or products, and then distributes that information with all the systems that feed the hub in real-time. A data warehouse, on the other hand, collects both shared business entities and the corresponding transaction history, such as customer sales orders, payments and service requests. Typically, a warehouse collects this data in batch mode and distributes the information to a select set of users via reports and graphs.

The enterprise data hub allows all corporate users to make real-time business decisions due to the active nature of the hub. The enterprise data hub improves any data warehouse solution, as data quality issues are more successfully resolved upstream, as an operational process, than downstream during the data warehouse ETL (extract/transform/load) process.

The Three Pillars of Enterprise Data Hubs

Understanding how the enterprise data hub model works starts with a look at its three core pillars: a master identity (one source of truth), active data quality services and continued synchronization with data sources.

Centralize The Data. Designing a data hub from scratch can be difficult, but there are out-of-the-box solutions that provide complete functionality. For evaluating options, there are practical considerations. First and foremost, the hub infrastructure must be anchored by an enterprise-level data model with the flexibility, power and scalability to manage the high-volume, 24/7 information demands of large corporations.

Apply Continued Data Quality Services. From an operational standpoint, it is cost-efficient to manage all data quality services at the enterprise level, instead of relying on departments to implement their own homegrown solutions. As data enters the hub from across the enterprise, duplicates are identified, accuracy is verified, source systems are cross-referenced and the blended records are assimilated. This approach frees individual systems from data management and eliminates hidden redundancies.

As such, the hub provides a unified view of each customer, company-wide. This helps to enforce global standardization, since all processes and validations now must align to the same data model.

To help ensure the accuracy and integrity of the customer's original data, some hub solutions offer real-time integration with third-party content providers. This enables the ability to purchase and download information directly into the hub so that a customer's business credentials, financials or credit/risk profile can be validated. Additionally, dynamic searches for matching companies can reveal unknown business relationships and corporate background, such as learning that a new customer was recently acquired by a corporation with whom it also does business. This functionality within the hub solution makes executives smarter and enables them to make better, more informed business decisions relating to their customers.

Synchronization. While everyone in the organization profits from a unified view of customers, each application system uses and needs different information. To avoid system overload, advanced hub solutions provide the means to control the flow of information to and from the hub. Using a sophisticated process control and publish/subscribe system, the hub can write updated records back to the appropriate applications, ensuring they operate with the latest, most accurate information.

The Investment that Scales

Enterprise data can only reveal so much. While sales numbers may tell how much business closed in Europe for the last fiscal quarter, they won't reveal that only 25 percent of deals closed, or that a major customer broke off negotiations and subsequently signed a contract with a competitor. Marketing totals also will not explain local purchasing trends, average lead generation costs or how effective a national marketing campaign was in targeting a specific audience.

These and countless other business answers can be extracted from a centralized data hub. With built-in analytic tools, managers can model customers in all their complexity and look at customer data consolidated from across all operations, rolled up into hierarchical structures (such as a family of companies or households), or graphically segmented for analysis. This real-time visibility allows for fact-based decisions to drive change and increase efficiency. The financial picture is dynamic, and can be adapted to reflect sudden market changes or windows of opportunity, such as spikes in supply or demand.

To be meaningful, customer data must be up-to-date and fit to be reported. With scores of data silos around the world, companies operate with huge information "blind spots." It can take weeks--even months--to collect, reconcile and analyze corporate performance data. And, at a time when corporate accountability is under intense scrutiny, corporate and financial officers need current, enterprise-wide business intelligence. This need is driving adoption of the enterprise data hub model.

RELATED ARTICLE: From Concept to Reality

The restaurant chain IHOP Corp. (International House of Pancakes) perfectly illustrates the on-the-ground financial potential of the enterprise data hub model. IHOP's Restaurant Support Center, the company's term for its corporate headquarters, realized that the organization lacked a view into its franchises and wanted better visibility into the day-to-day operations of each of its restaurants. Further, each franchise maintained its own data--legal, finance, operations, marketing, etc.--and had its own resources devoted to managing that data. As a result, business decisions were being made without all the data required for making the most accurate assessment. In short, executives and managers could not "vouch" for the very data they were relying on to make key business decisions.

IHOP has begun to deploy its version of a data hub--a franchise data hub--to connect all the different "silos" of information. In doing so, the company not only eliminated the need for multiple data managers but also ensured that every department could see every interaction that had taken place with every customer.

The second phase of the deployment will focus on gathering customer data and comment cards to better service its customers and improve targeted marketing campaigns.

This unified view will impact IHOP's operations at every level. At the service level, customer interactions will become more efficient. From a strategic standpoint, IHOP will be equipped with real-time access to critical data on customer spending habits and regional trends that it hasn't had before. The hub will also collect data from the restaurant's point-of-sale systems and various business applications (which will manage finance, human resources, etc.) to get a clear, accurate view of all the information that is coming out of each restaurant.

Bennet K. Yen is Senior Director, Oracle Applications Development. He can be reached at [e-mail].