The Next Commodity Layer of Business Intelligence

Traditional business intelligence models consist of several layers of technology, each one of which performs a key function in compiling, filtering, and presenting information to corporate users. The base layer consists of operational systems—such as legacy mainframe applications or e-commerce back-office services. The next layer contains a relational database engine—a data staging service, a replication service, or an information factory component. Then, to perform analysis, the system requires a transformational layer that filters and aggregates raw data into a timely, beneficial business view. Finally, visualization products—reporting and analysis tools—sort, merge, pivot, format, and distribute data according to users’ needs. Each one of these layers typically is comprised of commodity hardware and software components.

Business intelligence: a layered solution.

This “layers-of-commodities” approach generally is considered a best practice for creating a business intelligence infrastructure because it:

• Eliminates most of the software development lifecycle and its risks
• Encourages advantageous price competition among product vendors
• Enables optimization of product layers for specific performance criteria

But, there are some key drawbacks to this model. When it comes right down to it, a business intelligence solution is a composite of commodity software and locally customized—and often complex—data models.
Although industry-specific data mart design templates exist, no corporation is like any other. As a result, many business users are unable to navigate their custom data designs to effectively visualize the information they really need. Complex data marts may overwhelm users. Simple ones, on the other hand, may empower users to self-serve, but eventually users’ needs will evolve, requiring support from IT to produce the necessary reports.

Metadata: the icing on the cake.

The challenge is to create consumer-acceptable, consumer-empowering business intelligence technology. That becomes possible when you add a new business intelligence technology layer to existing systems: metadata.

Consumer information experience has changed radically in the past 10 years—in large part due to use of the Internet and search engines. Employing corporate business intelligence consumers with a search engine-interface to the business intelligence system provides them with a familiar and empowering way to answer the question, “Where can I find…?”

Metadata applications certainly exist; many are found embedded in market-leading transformation and visualization tools. These tools, however, tend to demonstrate a natural bias for their parent company’s business intelligence product line, and they tend to have limited change management and workflow solutions.

A true metadata business intelligence “layer” is an application customized for the ultimate user, providing a range of capabilities that can include:

• A comprehensive search engine
• Universal vendor architecture applicability
• Awareness of legacy system architectures, reporting system repositories, and workflows
• Pointers to owners, stewards, and subject matter experts

When the user is able to employ comprehensive self-help navigation tools as part of the business intelligence solution, the experience and impact increase in turn. And the metadata technology layer becomes the proverbial icing on the business intelligence layer cake.

West Monroe Partners works with companies to design and implement effective business intelligence solutions. For more information, please contact Dave Mausner, This e-mail address is being protected from spam bots, you need JavaScript enabled to view it .

In this issue:

• New Year's resolutions for 2008: Managing IT in the face of economic uncertainty
• Microsoft release casts new attention on CRM technology
• Greening the supply chain: Cutting costs while minimizing the environmental impact of businesses