InformationWeek

MIT's INFORMATIONWEEK 500 STUDY
The Productive Keep Producing
Successful companies support good business plans with the right information technologies

By Erik Brynjolfsson and Lorin Hitt
Issue date: Sept. 18, 1995

n the past couple of years, U.S. companies have gone on a computer spending bi nge. But is technology alone enough to ensure success? To ask the question is to answer it. Computer vendors can provide some of the ingredients needed to increase productivity and profits, but a cutting-edge computer shop that isn't aligned properly with the company's strategy and organizational structure is like a high-strung greyhound chasing a mechanical hare: No matter how fast the dog runs, it never reaches its goal.

The Hunt For IT Value
The leaders among the InformationWeek 500 are more than just big spenders. That's why we worked with InformationWeek to completely revamp the rankings for 1995. Instead of simply focusing on whether the companies in our survey use the latest technology, as past rankings did, this year we looked at their performance as well. The top organizations in our rankings have combined computers with new strategies and structures to generate more wealth than their competitors.

Focusing on performance instead of just technology is an important st ep forward, but we didn't stop there. After creating such a ranking, it is natural to ask what separates the top dogs from the mutts. To get some insight into this question this past spring, we worked with our students at the MIT Sloan School and examined six matched pairs of successful and unsuccessful users of IT. Our research suggested that certain business strategies, organizational structures, and information systems tactics can each play a role.

Then, working with InformationWeek , we polled both the chief information officers and human resources directors at hundreds of companies identified by Computer Intelligence as leading computer users. We collected data on their IS budgets; the extent of outsourcing and client-server systems; the use of electronic data interchange and other labor-saving technologies; the extent of business process redesign; and dozens of other aspects of the companies' business structure. The ones we found to be best at using IT to improve productivity are highlighted by a red dot in each of the 20 industry charts that accompany the sector stories on the following pages.

Granted, in a sample as large as ours, we found no single guaranteed path to success. But we were fascinated to discover that the aggressive use of computers is associated with a coherent new strategy of customer focus and worker empowerment. In contrast, a more hierarchical approach to management is more common among less-computerized companies. Let's look more closely at the patterns that emerged in each of the three areas we examined: strategy, structure, and IS tactics.

Customer Is King
Starting with the end in mind, we asked our respondents what strategic goals they had for their IS investments. Two broad strategies emerged: Some focused on costs savings and improved management control; others had a customer orientation, making investments in quality, customer service, flexibility, and speed.

Since our productivity analysis was based on "hard" numbers such as revenue, labor costs, and capital costs, you might expect that a hard-nosed, bottom-line orientation toward cost-cutting would do better than investing in "soft" benefits such as customer service. After all, customer loyalty may be a valuable long-term asset, but it doesn't show up on this year's balance sheet. But remarkably, the customer-oriented companies had the best productivity performance, outpacing the rest of the sample by a statistically significant margin (see chart) . What's more, when we looked at the profits of these companies, their winning strategy was equally apparent. We found no evidence that the companies using IT to save labor costs were as successful as their competitors at boosting productivity. If anything, they seemed to do slightly worse in both productivity and profitability.

CEOs and CIOs have been frustrated at the lack of accurate methods for evaluating IT investments. Consequently, many projects are j ustified mainly on the basis of cost savings, such as how many "heads" will be saved. Our analysis sounds a cautionary note for the simple expedient of a narrow focus on cost savings. Even when measured in productivity terms, this strategy may not deliver as much as considering the less-tangible benefits associated with a customer orientation.

Instead, more creativity is called for in thinking about new ways that computers can create value. General Motors spent billions of dollars automating its factories in the past 10 years to reduce labor costs, but, according to research by Haim Mendelson and Krishnan Anand at Stanford University, up to 20% of sales are lost each year. Potential customers don't find a car with the right combination of features on their local dealer's lot, and another 40% have to buy a "compromise vehicle."

Perhaps some of GM's IT budget would have been better spent on the kind of point-of-sale (POS) system that Kmart installed for $1 billion. The retailer links 2,250 stores by sat ellite to track sales and customer preferences daily. Or perhaps better information systems would make it easier to move to a just-in-time (JIT) inventory system that would bring GM's cars to the right dealers quickly.

The benefits of such systems can often be difficult to evaluate. But as our analysis shows, companies that pursue customer-focused strategies reap significant rewards.

New Work Structure
One problem with using technology to get closer to customers is that it can overwhelm decision-makers with information. After you collect gigabytes of data on customers by product line, location, date, other products purchased, etc., what do you do with it?

Retail is detail, and that means it often defies aggregation into a few broad statistical categories. Technologies like massively parallel processing can help identify patterns, but even some of MPP's biggest proponents are acutely aware of its limitations.

For instance, Wal-Mart uses a system much like Kmart's to gather data fr om each of its locations, but it doesn't try number-crunching its way to all the answers. Individual store managers retain responsibility for pricing on several hundred key items so they can match local competitors and respond to local trends. So while they use cutting-edge technology to stay close to the customer, their organizational structure has also enabled them to outpace the industry.

This pattern was reflected in InformationWeek's survey of human resources managers. The top companies in our sample tended to delegate more decision-making to line workers and said computerization was associated with skill-upgrading of their work force.

Shifting responsibility to the line has two benefits. First, it can help ease the information overload that makes the traditional hierarchy dysfunctional. In earlier eras, information was a scarce resource; now, it is managerial attention. Delegating responsibilities prevents the bottleneck at the top of many organizations that slows their information metabo lism.

Second, a strategy of true customer intimacy depends not only on hard numbers, but also on the kind of on-the-spot local information and intuition that defies computation and even communication. In such cases, moving decision-making responsibility to the field is better than trying to move information to the traditional decision-makers.

Henry Ford let his customers buy any color car they wanted, as long as it was black. His production and coordination technology made this the most profitable strategy available, and it didn't require him to know much about how his customers differed from one another. Also, if an employee approached him with a new idea, Ford knew whether it was likely to be a good or bad idea. Few of Ford's employees had any specialized knowledge that he didn't have himself.

By contrast, today's winning strategies use detailed information the boss can't possibly know. Not only must individual customer preferences be considered, but the technology of production also has risen t o a level of complexity that often requires more skill and authority in the work force.

More Than A Passing Fad?
This new structure of work runs counter to many of the principles and processes of traditional business design. So perhaps it is not surprising that we found that winning companies were more likely to be in the midst of reengineering or business process redesign (BPR) and devoting more of their IT budget to BPR than were their competitors. Many principles associated with BPR--such as reorganizing around customer-oriented processes instead of functions, heavy use of IT, and empowering workers to make decisions--are consistent with the new structure.

On the other hand, many companies turn to reengineering in response to a crisis in sales or earnings and, like medical care, most of the attention is lavished on patients who may be close to breathing their last.

What's worse, the term BPR is often used as a cover for downsizing and layoffs that involve no fundamental changes. Th us, one might expect that the "shadow of death" would dim the performance of companies that claimed to be aggressively reengineering.

The correlation we found between reengineering and business performance is thus more notable. Perhaps reengineering is more than a passing fad.

In contrast, companies that try to jump on the outsourcing bandwagon may be chasing a parked car. We found no association between outsourcing and success. If anything, companies that outsourced more of their IS work tend to have lower productivity and profitability.

The only performance measure heavy outsourcers did well on was stock market returns. Other research has found that, in the short term, the market reacts favorably to outsourcing announcements--and our analysis confirms this relationship. Whether the market will continue to react favorably, especially if productivity doesn't improve, remains to be seen.

Before IS managers breathe a sigh of relief at these findings, they should also know that the common alterna tive to outsourcing-- maintaining their traditional fiefdom of centralized IS spending--was not associated with significant productivity gains, either.

So what works? The data indicates that companies that spent more putting PCs in the hands of their employees and that moved aggressively toward client-server architecture outperform their rivals in productivity, profits, and stock returns. It's not surprising that these IS strategies are closely associated with decentralized decision-making.

When wild dogs go hunting, they face a choice: Work as a team and try to bag big game like a stag, or chase rabbits individually. Trying to catch a stag without a coordinated effort is likely to result in some painful, even fatal, antler wounds.

The survey results show a similar need for coordination among IS, line management, and overall business strategy. In our research, successful companies were more likely than their competitors to pursue the new structure of work based on decentralized decision-making an d increased work force skills.

Interestingly, another cluster of companies had some success with an alternative strategy of keeping decision-making centralized and using IT mainly for monitoring and routinizing work. What appeared to be unsuccessful was pursuing a mongrel of centralization without monitoring, or delegation without upskilling. The lesson: Go for the big game or stick to the old hunting grounds, but don't try to do both.

The risk of fence-sitting was illustrated very concretely at one of the companies we visited. The company undertook a major reorganization of its factory, substituting highly flexible, programmable machines for old dedicated machines.

While designing the new system, some of the design engineers decided to "rationalize" the product line and work flow, taking over decision-making from the line workers. While the line rationalization created efficiencies, it was only a quick win--a rabbit--and it undermined the broad coordinated effort to land the big game of a flexibl e factory with an empowered work force. With the benefit of hindsight, this quick win cost more than it gained.

In many ways, key lessons from this survey reinforce insights that academics and consultants have stressed for years: Get close to the customer, empower your workers, redesign business processes to take advantage of IT, and create a partnership among the IS, line, and management.

But these trends should not obscure the fact that many companies that ranked high on each practice still were not successful. Good execution of coherent policies is what matters. This requires an ear for customers' desires and a nose for technology's capabilities that can't be reduced to a financial statement or a checklist.

The race does not always go to the swift, but with some good coaching they do have an edge.

Illustration for InformationWeek By Kathleen McCutcheon

Comments on this story?

InformationWeek http://techweb.cmp.com/iwk