Global/Multicultural
References Work Place Case
August 15, 1994
Vendors are partly to blame. Even though we are now learning how to work around the classic problems of inflexible old mainframe "legacy systems," some of the newer technologies are still not delivering their promised benefits.
For example, consider the trend to install ever more complex personal computers in the offices of white-collar workers. PC users waste, according to one consultant's estimates, 5.1 hours each week "futzing" with their computers-learning how to use them, waiting for them to do things. checking the things they do and so on. And that doesn't even measure the time wasted by employees playing games loaded on their PCs as standard equipment. If these numbers are correct, then PCs have become the biggest destroyer of white-collar productivity since the management meeting was invented.
However, the real problem lies at the border between human skills and information technology. For example, one bank started to roll out complex PC-based systems using a sophisticated graphical user interface to encourage its branch management to sell more. Despite the considerable costs of this program, sales remained flat. Not only were the managers already overloaded, but they had neither the skills nor the motivation to become salesmen.
By contrast, another bank built all its selling skills into simple-to-use technology. The bank then assembled a new, lower-skilled work force to "do what the systems told them." Sales force productivity tripled. In a growing number of industries, the use of "intelligent" technology— together with a retooling of human skills around its potential-can help to radically reshape business strategies in a remarkably short period of time.
In the automobile insurance market, ADP, a specialist financial services provider, has recently developed a system based on portable computing that can allow an assessor to give an on-the-spot calculation of insurance claims. A pen-based computer carries in its CD-ROM storage a data base of tlle parts for every model of car routinely found on U.S. roads. It also has enough machine intelligence to enable a relatively unskilled operator to come up with an immediate estimate of repair costs.
This information is then transmitted, using a digital image of the accident if necessary, to the insurance underwriter. Using sophisticated interconnect technology, the information is then passed to all relevant parties.
Viewed solely as an improved processing system, this technology will have clear efficiency advantages, but its real benefit will be in enabling radically improved customer service. Why, for example, should customers have to submit complex claim forms and then wait for their money? After all, they had to pay the premium up front. Insurers, using these IT systems, can now offer a "pay at the pump service," which hands the accident victim a check on the spot.
Corporations who use this type of IT can also learn more about their customers; this can then be exploited with an expanded product line. The expertise with which this knowledge of customers is nurtured and managed then becomes a powerful competitive weapon to strengthen relationships and build sales, and is ultimately the key source of competitive advantage.
For example, Fingerhut, a Minneapolis-based mail-order company, had to build sophisticated credit evaluation systems to manage its low-income customer base. However, it was then smart enough to recognize that credit data are also a rich source of behavioral information, especially when augmented by analysis of the actual purchases made by each customer. The company therefore developed its own customer-management system, which now contains more than 1,000 items of behavioral information per household and allows exact targeting of sales campaigns. Its resulting financial performance has been exceptional.
Given the potential of these new technologies, how then can companies developthem without the agonizingly slow design and build cycles that have characterized most experiences with new systems main problem may be in trying everything right the first time intensive, monolithic software development projects. T. Capers Jones, a Burlir Mass.-based expert on software development, has calculated that more of what large, successful computer systems end up doing is developed after the system goes live.
The greatest degree of software innovation occurs in a "living laboratory" where flexible developers and entrepeneurial users work in partnership. Such combinations of focused brain power may take place internally or in specialist software providers, which are now springing up all over the world of intelligent technology.
Increasingly, new processes can therefore be developed by specialists into software packages. They can then be "shrink wrapped" by intelligent technology so they can fit immediately into a new operation, without major delays. The auto insurance estimation tool, with its short learning cycle and sophisticated interconnect facilities, is obviously an example of such process packaging. Using such built-in capabilities, wholly new businesses can now be constructed in record time.
Since innovation cycles that products such as personal computers obsolete in 18 months or less can now affect whole corporations, developing a flexible and incremental approach may well mean challenging some of the basic theory we have used in the past. Long cycles for IT development must be replaced by new businesses and services based around the commercialization of existing innovation. Human resource planning also need to be revised to cope with a world where the role of general management hierarchies may virtually disappear.
Barriers to exploiting the potential of intelligent technologies are frustrating. But if managers ignore the experierence of the entrepreneurs, then their companies will be left tinkering with legacy systems while new players steal their customers and markets.
Mr. Heygate is a principal in McKinsey's London office.