Why do so many corporate investments in information technology fail to deliver their full business benefits? A recent yearlong research study from the Accenture Institute for Strategic Change, entitled "Data to Knowledge to Results: Building an Analytic Capability," concludes that organizations don't pay enough attention to the creation of the analytic capabilities to be delivered by that technology. In other words, the systems are doing just what we asked them to do – delivering optimized transaction processing – but we're not asking them to do enough. Only the right analytic capabilities enable an organization's people to draw useful insights from the data – insights that improve the bottom line or fulfill a particular mission.
If corporations are to realize a return on the millions – even billions – of dollars they've invested in new systems for the back and front office, they need to deliver the real data and information necessary to manage the business. One of the key steps in delivering that capability is an enterprise data strategy.
What is an Enterprise Data Strategy?
An enterprise data strategy is a plan for improving the way an enterprise leverages its data, allowing the company to turn data into information and knowledge which, in turn, produces measurable improvements in business performance. For commercial enterprises, the improvement is in increased revenues or decreased costs. For government agencies and not-for-profits, the improvement is in the way the organization fulfills its mission of service to its patrons.
The enterprise data strategy works at a high level and with broad brush strokes; however, it's a critical aspect of an overall approach to strategic information technology. The strategy involves four components: an enterprise data architecture (see last month's column for more on data architectures), data management, enterprise analytic capabilities and an action plan. With the data strategy in place, the action plan drives each of the other three components to be effective alone and in their interactions with each other. (See Figure 1.)
Figure 1: Enterprise Data Strategy
I begin with the analytic capability here to avoid the mistake many enterprises make: beginning with the technology and then thinking of ways to use it. Instead (legacy systems notwithstanding), try to begin with what you want your people, your business partners and your customers to do with the technology. There are many kinds of analytic capabilities, of course. Customer relationship management programs, for example, are bringing more powerful analytics to the analysis of customer behaviors and preferences, resulting in the development of better products and more effective marketing programs. Supply chain analytics are improving the efficiency and effectiveness of manufacturing and inventory management. Financial analytics are improving allocation of financial resources to increase overall returns on investments. All these types of analytics should be considered as part of the overall data strategy.
As I discussed in last month's column, an enterprise data architecture includes four elements: organization or structure of the data, storage, access and the movement of data to locations where it best serves an organization's people. Usually, only two of these are addressed in an enterprise data strategy: the structure of the data and its movement across the enterprise.
Typical strategies for structuring data include the use of data warehouses, operational data stores, data marts to support analytics, interactive databases to support dynamic interchange with customers and suppliers, and traditional back-office databases.
Typical strategies for moving data across the enterprise increasingly rely on enterprise application integration (EAI) technologies to provide real-time delivery of transaction data to the multiple parts of the enterprise that need it. However, difficult issues arise with older systems not designed for real- time interactions. For that reason, the strategy must address plans for dealing with legacy systems as well as newer, innovative systems.
Data management refers to the traditional requirements regarding the care and feeding of production databases. It also encompasses the management of data models to guide the design and building of the data architecture and to identify data redundancies across the enterprise. Those redundancies can be addressed through the use of EAI technology to improve synchronization. However, these traditional areas of data management do little to help the enterprise turn data into knowledge and then into business results. New data management components include comprehensive data quality programs and data guide programs.
Data quality programs recognize that decisions are only as good as the data on which they are based. It follows, then, that improving results requires improving the quality of data.
Data guides are individuals who provide the person-to-person support and mentoring needed so that employees can use relevant data in the organization to create knowledge and produce results for the enterprise.
Significance for Today
In today's economy, where companies are struggling to maintain revenue and are tightening their budget belts, an effective enterprise data strategy can be a critical part of creating new opportunities for business growth by leveraging available data. A data strategy that ensures that information is translated into optimized performance can improve revenues, reduce costs, improve resource allocations and deliver greater value to stakeholders.
Next month's topic will be: Creating the Action Plan.
Nancy K. Mullen is retired from Accenture's Data Architecture specialty. Mullen is a frequent speaker on the topics of architecture and database management and coauthor of NetCentric and Client/Server Computing: A Practical Guide (Auerbach, 1998), focusing on data warehousing. She may be contacted at NKMullen@aol.com.