|
Executive Overview
Today's business intelligence (BI) software products are expected to handle a broad spectrum of activities, from scheduled reporting to ad hoc queries to complex analytical processing. To meet user needs, BI vendors that were formerly concerned mainly with enterprise reporting are now delivering enterprise business intelligence suites (EBIS) consisting of production reporting, portal connections, spreadsheet connections, ad hoc reporting, and OLAP viewing. This makes it easier to purchase and maintain software assets because IT managers can minimize the number of vendors they work with. But it also necessitates careful planning and cost projections to ensure IT resources aren't overwhelmed by burgeoning use of the tools.
"With the move toward convergence, enterprises can begin planning to reduce or limit the number of vendors and technologies they use for many business intelligence (BI) activities," says Gartner analyst Howard Dresner. "This suggests a needs assessment process to determine how BI technologies are used today and how they will be used into the future."
Business intelligence tools were initially dominated by analysts and power users trained to use multidimensional databases, online analytical processing (OLAP) tools, and other specialized software. The new breed of business intelligence applications enables casual business users to access and analyze information through intuitive user interfaces using little more than a Web browser. As BI tools become more capable and widely used, organizations must be able to accurately gauge the associated costs over both the short and long term. IT managers must consider not only the up-front costs for hardware, software, and infrastructure, but the ongoing costs of maintaining and upgrading those solutions over time.
Yet many organizations roll out BI solutions in an iterative fashion without giving adequate consideration to the optimum way to deploy, upgrade, and maintain their software assets. In short, they fail to consider the total cost of ownership (TCO) for these important solutions. "Organizations often focus on visible cost drivers like product-license costs and consulting fees," suggests Jim Mullen, business intelligence practice leader in Information Builders' consulting organization. "But it is more difficult for a business to forecast the total expected breakdown of maintenance costs ?labor and assets ? for any given IT solution."
Getting a Handle on BI Expenditures
There are dozens of cost elements to consider when calculating TCO, most of which can be grouped into three major categories: purchase costs, implementation costs, and maintenance costs. Some questions to ask at the outset of a BI implementation include the following:
- Which reports require real-time information versus data from a warehouse, and how will the infrastructure accommodate them?
- Can the BI architecture be economically expanded to support a growing user base?
- Can the tools handle constantly expanding internal user requirements as well as external deployment to customers, partners, and other third parties via the Internet?
- How much does the BI vendor charge in license fees for self-service users (who are not named or identifiable)? Do they make a licensing distinction between generalized reporting and personalized reporting?
Answering these questions requires a solid understanding of the BI architecture ? how it is configured, how it is deployed, and how it is scaled ? as well as an informed estimate of the current and potential user base. Armed with this information, you can predict how deploying specific types of BI applications will impact costs ? both the cost of various types of user licenses and the hidden costs associated with expanding the BI infrastructure. It is particularly important to know precisely how a vendor will charge for each type of user, since it can have a big impact on overall costs down the road.
|
