From...

June 23, 1999
Web posted at: 8:58 a.m. EDT (1258 GMT)

by Leonard M. Fuld and Kent Potter

(IDG) -- Corporate librarians know the score, so when we wanted to know how effectively major companies use the information they buy, we talked to librarians at seven large businesses.

In their opinions, companies fail to use as much as 70 percent of all the data they purchase. Run this estimate against the nearly $30 billion of Web and online services that research company Simba Information Inc. in Stamford, Conn., predicts corporations will purchase in 1999, and you get $21 billion. While the poll and its numbers are admittedly informal, the issue raised is relevant: Companies waste a lot of expensive data each year.

MORE COMPUTING INTELLIGENCE
	IDG.net home page
CIO home page
Make your PC work harder with these tips
Reviews & in-depth info at IDG.net
IDG.net's personal news page
IDG.net's products pages
Year 2000 World
Questions about computers? Let IDG.net's editors help you
Subscribe to IDG.net's free daily newsletters
Search IDG.net in 12 languages
News Radio
	CIO radio
	Fusion audio primers
	Computerworld Minute

What is data waste? We define it as the amount of online data a company purchases minus the data actually used by that corporation. Organizations often buy numerous online data services they do not need or purchase services with overlapping content. If a large corporation contracts for approximately $10 million of enterprise-wide online services, our librarians would estimate that it could be wasting as much as $7 million of the total.

If you're looking for a common example of data waste, check out your company's Intranet. Does its home page feature a news stream that races along the upper end of the desktop screen? These tickers create almost no decision-producing intelligence -- in other words, data waste. And data waste can turn into data overload when a manager can't find critical information in the sea of irrelevant data flooding the corporate networks. Once workers start wading through useless information or hunting for lost data, it's a short step to decreased worker productivity.

"Wasted data certainly has an impact on productivity and overall expense," says Herbert M. Baum, president and COO of Pawtucket, R.I.-based Hasbro Inc., one of the nation's leading toy manufacturers. Part of the reason for data waste, according to Baum, is its source. "I have found that when you buy data from an outside source, a lot of it is irrelevant," he says.

Data from outside sources, however, isn't inherently irrelevant. It's just that many companies are buying too many, or badly chosen, data services.

The causes of data waste

As with any sort of physical pollution, companies must attempt to control data waste. In order to do so, managers should root out the causes, including the following:

• Many executives ignore purchased online data, thinking it's not current. And since other companies have access to the same information, management sees little competitive advantage in using such data.

• Volatile business markets can cause a company to react and buy even more data rather than use the data in its existing data pool. This is a particular problem in fast-moving industries such as telecommunications, pharmaceutical and financial services. Corporate managers latch onto myriad new data sources in hopes of learning something new that will help them stay ahead of their rivals, more effectively support customers or merely survive.

• Companies tend to toss their procurement strategies to the wind when it comes to online data. Many of our clients have few, if any, central rules for buying online data. Departments often buy data separately instead of forming buying coalitions and leveraging corporate purchases for enterprise-wide discounts.

• Everybody wants the latest and greatest information, whether they need it or not. In one instance, a major corporation even experienced a data waste meltdown. The company, a business unit in a financial services firm, purchased an online push technology news service. The service offered limited filtering capabilities that forced everybody to receive almost the entire news feed. Nearly everyone signed on, and volumes of mismatched data poured into the firm. Hundreds of individuals at the firm received copies of the same news items, all in full text. Just a few weeks after installing the service, the company's servers froze from too much of the wrong data.

Minimizing data waste

Just as any executive worth his wingtips will have a strategy for keeping on budget, you can plan to buy data frugally and wisely. With a little analysis and up-front work, the following tips can save money and time.

Buy smart: Every large corporation should have an integrated buying strategy for online data. Somebody, whether it be you or a corporate librarian or the purchasing manager, needs to assemble all potential user groups, determine who needs to know what and tailor unique packets of data to suit the needs of each group. Include your firm's librarians on this buying team. Since many of the online services offer similar news content or repackage the content from common sources, you need these research experts to analyze provider content and help the procurement folks parse each service into clear, distinct categories.

Map data: Data mapping is the ability to match specific strategic and tactical business needs with the sources themselves. By mapping data, you'll be able to pinpoint which data sources are either too general or miss your organization's strategic and tactical needs altogether.

Data mapping demands that you first find out what your audience needs. In the case of one client, we interviewed more than 20 senior managers and discovered that nearly every person had the same five information needs in common. For example, almost all the managers wanted competitive intelligence on the new financial services technology being delivered by the company's competitors.

But when we explored the online and intranet-available data the company had purchased, we discovered that most of the sources proved totally irrelevant to those five information goals. The needs assessment then allowed us to map to other online sources—some even free on the Net—that better matched their day-to-day activities.

Know the difference between relevant and significant Relevancy doesn't reduce data waste, significance does. Modern search engines and the latest in artificial intelligence database-mining software do a remarkable job identifying relevant data. These packages will tell you how close your hits resemble the search terms you placed in the system. But software engines do little to determine how significant the findings are to your ultimate decision. Do not expect a software package to solve the data waste problem. At best, software reveals the relevance of a data pool, not its significance. Attempting to ingest mountains of relevant but insignificant syndicated research and news may only reinforce hip-shot analysis for crucial decisions, as anxious analysts snatch large stacks of available information to confirm their biases. Using piles of such data is not the way to shape markets as a leader. It is a way to bog down decision-makers.

Avoid using old data for new markets: Companies need to rebuild their data pools for today's market, ideally doing so every five years or less. But most corporations rely on obsolete data that's nearly useless for the new marketplace.

BellSouth Corp. in Atlanta confronted such a situation when it faced the proliferating competition of nontraditional telecommunications networks such as wireless, cable and Internet firms. New rivals brought with them new competitive strategies and a demand for wholly different pieces of market data. BellSouth widened its data net, using data from interviews with distributors, suppliers, industry trade associations and other local technology sources.

In Brazil, for example, BellSouth International gained significant market share in the cellular market by using Brazilian information sources that lay outside the traditional telecommunications data boundaries.

Data waste is not necessarily a new problem, but it is made urgent by the accelerating pace of the Information Age. Before your company spends millions on electronic information services, ask yourself this: How much -- and what kind of -- data does your company need? Eliminate the rest -- fast.

Leonard M. Fuld is president and founder of Fuld & Co. Inc., a Cambridge, Mass.-based competitive intelligence consulting and research firm. Kent Potter is vice president of intelligence processes and systems at the company. They can be reached through www.fuld.com.