Why does the stock market routinely value software firms at 25 to 75 times earnings? I can think of two answers, one from the supply side and one from the demand side. Together, the answers show how every company can use its information assets to capture similar valuations, at least for some parts of their businesses.

The supply-side answer is that, once software is designed and debugged, the cost of producing the "next" set of diskettes is almost zero. And if the software is distributed over the public telecommunications network, the cost is zero.

The demand-side answer has to do with the enticing mathematics of networks. Think Metcalfe's Law, which says that the value of a network grows as the square of the number of users connected into it. One phone does nobody any good, but 200 million become unbelievably powerful. [See the interview with Bob Metcalfe, "Law and Disorder," Page 38.] So, networks have incredible ability to keep expanding, and as they do they will create a reachable market for software that becomes ever more enormous. In other words, software offers tremendous economies of scope. The only limit is in how we imagine the opportunities for software.

Now let's generalize. Not just software, but any form of digitized data, in theory offers comparable economies of production scale and market scope. Unlimited supply, endlessly reconceivable demand.

This is why Federal Express has started referring to itself as a "logistics" company. It sees more profit from knowing how things move from one place to another than from actually moving them.

Let's back up. All companies understand the need for information. Almost all know they need to use technology to manage the information efficiently. The problem is that, historically, information counted for little beside the value of rolling stock, the oil oozing up from the ground, or the blast furnaces burning bright in the night. So, companies tended to view information only as a way to control their operations more tightly. But, by now, companies have amassed monumental amounts of information that could be used in other ways, producing the same kinds of magical economics that drive the software business.

To achieve those economics, companies need to do three things. They must determine the quality of the information they have. They must find ways to improve that quality. They must find novel ways to take advantage of unused or undervalued pockets of information.

LEARNING WHAT YOU HAVE. Surprisingly, even this first step isn't one that many executives have taken. How many executives know the accuracy of their customer database? In my experience, not many. When Travelers and Citicorp announced plans to merge, Citicorp's chief executive, John Reed, announced that his target for the year 2005 is a billion customers. I bet no one at Citicorp can do more than guess at how many customers the bank has today.

To take that first step, you need to think about your information assets along at least seven dimensions:

Reach—How far up and down the chain can you "see?" Into your suppliers' information systems? Into your retailers'? Beyond?

Depth—How deeply do you understand how one event causes another? In automotive engineering, information is deep. In advertising, it's shallow.

Timeliness—How fast do you gather information—and how fast should you? For the average guy dabbling in stocks, yesterday's quotes in The Wall Street Journal are timely. For a huge institutional investor, lagging 20 minutes behind trading is unacceptable.

Accuracy—Again, this is a matter of use. Instruments used in producing semiconductors must stay within a tolerance of 0.000001%. On the other hand, in a direct mail database, some names and addresses will always be out of date.

Fidelity—Do the measurements gauge what they claim to? Joe Pine, a specialist in mass customization, says customer satisfaction surveys are low-fidelity because they ask questions about what a company makes, not what the customer wants.

Granularity—What is the level of detail? Federal Express has highly granular systems; it can follow a package at every step along its route.

Flexibility—This is the interesting dimension. How easily can you shift views? Phone companies and auto makers have a wealth of data, but it's locked in static configurations. It's a nightmare to attempt to analyze telephone calling patterns or individual customer usage patterns.

IMPROVING THE QUALITY OF YOUR INFORMATION. This is partly just a matter of going through the seven characteristics of information and looking for opportunity. Maybe you can expand the reach of your information by making a data-sharing deal with your suppliers. Maybe you need to improve the flexibility of your information by adding the ability to organize it by customer, rather than just by product line.

But you probably also need to take a hard look at your accounting and incentive systems. You may have lots of information assets, but you probably don't carry them on your books, even if you're a credit bureau. Your computers show up as assets even though they can be replaced relatively easily, but not the data that provide your lifeblood. Why is that? In fact, quantifying the value of information is a hard problem because it depends so much on the context—knowing tomorrow's winning lottery number can be worth millions of dollars; knowing yesterday's is worth nothing. But we all need to start working on ways to quantify the value. Otherwise, we'll be stuck with today's accounting systems, which reward executives for using information efficiently but provide little incentive for managers to gather information that someone else in the organization might be able to use.

FINDING NEW USES FOR INFORMATION. This is something that some companies are, in fact, attempting. Some retailers are giving suppliers access to timely information on sales and inventory levels. All sorts of companies are trying to emulate Federal Express and give customers access to what had been internal information, about delivery schedules, manufacturing schedules, etc. Some competitors, such as catalog retailers, even find ways to aggregate information about customer purchases, so they all know which customers to solicit heavily.

These plans to share previously unused information can increase revenue by helping identify strong prospects, or can cement loyalty among partners or customers.

Opportunities will become more pronounced over time as technology standards make it easier to share information among companies with different systems and different approaches to data. But, executives still have to have the discipline to look for opportunities. In the future, how you manage your information assets will matter much more than how you handle the kinds of physical assets currently on your books.