With so many companies rushing to plunk down tens of millions of dollars to install customer-relationship-management systems, not enough companies are stopping to think first about how to get the most out of the software. That helps explain why a study found that more than half of CRM installations in 2001 were not meeting expectations.

The theory is that CRM will let companies better interact with customers, track their behavior, squeeze out internal inefficiencies, raise customer satisfaction, and improve customer retention. The problem is that companies mostly are tracking just what they’ve always recorded about dealings with customers, rather than seizing a golden opportunity to rethink what they want to know.

According to local legend, Boston’s streets are so notoriously hard to navigate because early residents merely paved over existing paths, forcing today’s cars to live by the dictates of the 17th century. Similarly, many companies are using CRM to merely “pave the cowpaths,” to use the old I/T expression.

Instead, companies need to take the time to design the equivalent of a modern road system. You need to do the upfront work to make sure that you know just what problem your CRM “solution” will solve.

There is no “silver bullet.” No generic CRM application is going to tell you just what to measure. So you need to figure out how to answer at least six questions:

What do we want to know about current and prospective customers?

What is this knowledge worth to us?

Who needs to do what with this knowledge and when?

What should our incentive structure look like?

How do we know if we’re doing the right things?

Do we have the appropriate technology infrastructure? If not, what will it take to get it?

Once you have the answers, here are some steps you’ll likely need to take:

RESET YOUR SIGHTS. In the paper-packaging industry, some companies have found purchasers of specialty products to be more desirable customers than those buying standardized cartons. That’s because buyers who spend lots of time collaborating with their suppliers are more likely to spend more money and remain loyal customers.

As a result, these packaging companies can de-emphasize traditional measures of profitability, which tend to provide a snapshot of what happened weeks or months in the past.

Instead, the companies can construct measurements that will not only provide a better snapshot of where a company is today, but will also give the company a chance to quickly change its future. Those measurements might cover which salesmen are actually using their design departments to bolster sales.

LEARN WHERE THE BEST INFORMATION IS HELD. Companies have lots of information that can be useful—but only if it is made available to the right people.

At a food-services company where I once worked as a consultant, I got a bird’s eye view of the problems that come from lacking key information. Management had provided incentives for salesmen to sell goods that had high profit margins. As a result, the sales force pushed paper products, such as napkins, rather than luxury goods, such as lobsters, which carried far smaller markups. The problem is that the profit margins didn’t factor in shipping. Paper goods are far more expensive to ship than lobster because, per dollar of sale price, they take up so much more space in a truck. Because management didn’t understand what its most profitable products were, it struck deals to sell and ship lots of marginally profitable or unprofitable items.

Once the company switched to the right measure—profit per cubic foot of truck space—and continuously communicated that measure to its sales force, the problem began to go away. The company had put the right information in the hands of the right people, and they could act on it.

CREATE THE INCENTIVES TO MAKE SURE CHANGES BECOME PERMANENT. I recently ran across a sales force whose incentives were tied to the percentage of gross margin on goods sold. This compensation structure was flawed because it inadvertently encouraged the sales force to mark up prices to levels that drove customers away.

A better idea is to tie compensation to something that more closely matches the overall company’s goals, such as retaining customers.

It’s also important to give people the incentives to ensure that the data collected are used and used regularly.

The process of converting information into profits drives modern organizations just as engines converted fuel into work in the Industrial Age. Now, instead of drilling, refining, and distributing fuels such as coal and oil, we need to capture, manipulate, and transmit data. Once the data are turned into information, they can be converted into something valuable.

We have a long way to go before we understand data as well as we do engines and combustion, but if we start to ask specific and pointed questions about what we need from information we’ll be moving in the right direction.

Munck is a managing director at DiamondCluster International Inc. He can be reached at peer.munck@diamondcluster.com.