Given this diversity of data sources, executives complain that they have no single source of truth for comprehensive data they can trust. They have tried consolidating data sources, but find that it only helps solve part of the problem and is far from a complete solution. Consolidating data typically involves taking data from operational systems and moving it or replicating it into a central location like a data warehouse. Such a project often requires 12 to 18 months for implementation and is increasingly difficult to modify as the warehouse evolves. More importantly, this approach doesn't allow real-time visibility of the data, which is changing every moment of the day.

Companies have tried buying large, monolithic applications to act as a single system of record, but still find that important data continues to exist on multiple systems and in different forms. Enterprises have become savvy enough to know that no one system can achieve all of their objectives. Best-of-breed applications will always have features and functions that are superior in some but not all aspects. In addition, companies don't want to be tied down to one system or vendor. Finally, purging legacy systems altogether is impossible. Previously deployed systems will always exist, in one form or another.

Is it practical or feasible to have a single source of truth? We propose that the most viable, pragmatic solution is a composite view of record (see exhibit 1). Just as a database view is built to abstract multiple tables in a single data source, a composite view is built from multiple underlying sources. Composite views are created based on software technology called enterprise information integration (EII). EII technology, as an emerging enterprise software category, is capturing attention because it addresses many of the issues just described. Although similar to relational database views, a composite view provides a higher-level abstraction layer, spans a much wider range of data types, and can represent data stored in multiple locations and varying formats.

The benefits of this new layer can be best illustrated by considering a classic example -- the need of a marketing organization to have a single view of the customer. Today, businesses are finally able to shift away from a myopic focus on the bottom line. The beginning stages of a recovering economy have turned their goals instead toward increasing top-line revenues. Marketing teams, as a result, are being scrutinized for their ability to design and implement promotional and advertising campaigns that increase sales to existing customers, reach new customers in existing markets, and penetrate new markets. A marketing view of record (see exhibit 2) can be constructed to aggregate information from CRM software, financial applications, customer information systems, and product information. Attempts to assemble data from each of these sources would ordinarily put the marketing team at the mercy of scarce IT programming resources every time it wanted to query one of the company's many databases. But a composite view can be built interactively, providing a real-time single source of all information and allowing marketing teams to be less dependent on IT for querying this data. Similarly, the results of marketing campaigns can be effectively assessed using composite views of record. Users may also want a quick picture of individual customer activity across product lines, or as it relates to the channels used for a particular campaign. To do so, they simply create a view composed of the relevant product lines and channels. Instead of looking in multiple places to patch together this information, they simply go to the view.

HR departments can also benefit from composite views. Managers can use a composite view of each employee to evaluate and monitor compensation packages that include salary, commissions, and benefits. Each employee's history can be viewed collectively, even if they've moved to another site or transferred between different functional areas.

As a single source of truth, composite views ensure access to accurate, relevant data. The relevancy issue has presented considerable challenges to businesses in the past. Decision-makers in today's information-intensive businesses are often unaware of all the relevant sources of information that pertain to a particular decision or task, or cannot easily gather all of the relevant data into an understandable "big picture" of the issue. In many cases, they're inundated with too much information, which can contain stale or unrelated data, and the relevant facts aren't distinguishable. Software vendors and technical and business analysts can help companies develop composite views that consolidate the data users need but spare them from information they don't want.