What it Takes toBy Allan Alter | Posted 10-05-2005
Business Intelligence Is Valuable, but Falls Short of Its Potential
Ever since 1964, when a researcher named Michael S. Scott Morton coined the phrase "decision support system" in a basement office at Harvard Business School's Sherman Hall, technologists have tried to build systems that can help managers understand what's going on inside and outside their companies, and thus make better tactical and strategic business decisions.
Scott Morton went on to a successful career as a professor at MIT's Sloan School of Management, but his brainchildren have had spottier records.
From the executive information systems that sprouted up in the 1970s to the expert systems of the 1980s and the knowledge management systems of the 1990s, companies have struggled to pull together large quantities of data, ensure its accuracy, write rules and algorithms that created usable information, and deliver that information in an easy-to-understand form.
Their brethren business intelligence systems first arose in the early 1990s as a way for businesspeople to bypass the IT organization and get information directly from data warehouses or data marts, and they too ran into trouble in the beginning.
But in the past five years, BIa term used to describe everything from business performance management to competitive intelligence to customer analyticshas become easier to use and is now widely supported by many IT departments.
Is the old dream of making better decisions through information technology coming closer to reality today? The answer is yes, judging by the results of this month's survey.
When four out of five BI users say these systems have had a measurable positive impact on the bottom line, and nearly 100 percent say they are critical to their business strategy, BI has to be considered a success, if a qualified one.
Business intelligence has "absolutely" improved business value, says Lillian Sullivan, director of enterprise information and analysis for LexisNexis Group, a $2.1 billion provider of legal, news and business information based in New York City. "It's the lifeblood of the business. It's central to everything our internal customers do."
The company, a division of media giant Reed Elsevier, has documented more than $40 million in additional revenue and cost savings since 2000 by using BI to improve marketing and customer service. Lexis-Nexis is using BI for customer segmentation, identifying customers who have yet to purchase services that others in the same group have bought, and targeting marketing campaigns to them.
Intelligence software, however, is really only as good as the people who use it. In another encouraging sign of its success, more and more managers and employees are using BI. The percentage of executives who use these tools when their companies provide them is higher than in our 2003 BI survey. What's more, BI is now used not only by top executives but by mid-level managers and planners at most companies that have installed the technology.
One-third of BI users have gone further; they have made BI an everyday tool for all kinds of employees. The staff in the purchasing department at Quaker Chemical Corp., a $400 million specialty chemicals manufacturer in Conshohocken, Pa., use BI to analyze the price of raw materials worldwide, while the collections department relies on the system to identify past due accounts and analyze their history with Quaker.
By showing users how much easier it is to use the BI tool than to navigate the ERP system they had relied on, "we are seeing the usage of our system grow 300 percent in the last two years alone across all metrics: number of users, number of times each user is using each system," says Vice President and CIO Irving Tyler. Satisfaction with BI seems to follow a corollary of Metcalfe's Law:
much as a telephone or computer network increases in value as more people join it, the higher the percentage of people at a company who use business intelligence, the happier they are with the system.
So the question today is not whether these systems are delivering value, but how to get them to deliver more. There's room for improvement: A third of our respondents still are not satisfied with the information they are getting, a figure that has not improved since our last BI survey two years ago. And while most respondents find business intelligence to be effective for getting insight into internal operations and improving financial performance, only about half have found BI to be effective at increasing revenues by optimizing pricing, identifying business opportunities or cross-selling.Some Specifics:
What it Takes to
What will it take to improve? Ensuring data quality is a necessity. And aligning systems with strategy shouldn't be overlooked; it's the most effective way to improve the value of BI, according to our respondents. Defining intelligence goalsa step fewer than half of our respondents are good or excellent atseems a necessary start.
But increasing the value of BI will take yet more effort. Thomas Davenport and the other authors of a May 2005 study of 32 organizations that enjoy strong business performance and success with business analytics learned that these companies "go well beyond basic statistics" and use predictive modeling and other sophisticated analytical techniques.
If more companies are to have success using BI to increase sales, says Gartner Inc.'s Simon Hayward, they "need some theory or model of why people behave a certain way, so you know what to do with the data. If I lower the price of A, will I sell more B? Or is it the other way around?" CIOs must find a way to make BI both easy to use and understandable, even while providing sophisticated capabilitiesand the know-how to use these decision-making tools intelligently.Scouting Report on BI in Corporate America:
Over 80 percent of companies that use business intelligence systems are generating business value from them.
Companies already using business intelligence are expanding their investment.
Satisfaction with BI information has not improved in the past two years.
BI's record for boosting sales is mixed.
Alignment with strategy and with user needs is the best way to get more bang from the business intelligence buck.
Expect trouble if data quality and integration problems dog your BI systems.
Business user satisfaction requires help with collaboration and interpretation.
Despite advances in displaying BI data, users still love to use spreadsheets.