Drucker on Management
Drucker on Management: Permanent Cost Cutting
Drucker on Management: Permanent Cost Cutting
By Peter Drucker
The Wall Street Journal - Jan. 11, 1991
Scores of large organizations -- businesses of all kinds but also government agencies, hospitals and universities -- have sharply cut staffs these past few years. But few have realized the expected cost savings. In some cases costs have even gone up. In many more performance has suffered. And there are growing employee complaints about stress and work loads.
Cutting staffs to cut costs is putting the cart before the horse. The only way to bring costs down is to restructure the work. This will then result in reducing the number of people needed to do the job, and far more drastically than even the most radical staff cutbacks could possibly do. Indeed, a cost crunch should always be used as an opportunity to re-think and to re-design operations.
To start cost-cutting managements usually ask: "How can we make this operation more efficient"? It is the wrong question. The question should be: "Would the roof cave in if we stopped doing this work altogether?" And if the answer is "probably not," one eliminates the operation. This is unpopular, to be sure. Some one is bound to argue: "We needed this procedure only 18 months ago and may need it again 18 months hence." But eliminating an entire operation is by far the most effective way to cut costs, and the only one likely to produce by itself permanent cost savings. It is by no means accident that the only places where cost-cutting done during the past few years has produced real savings are where an entire operation was eliminated -- in the commercial bank, for instance, that closed down an unprofitable merchant-banking subsidiary.
And it is always amazing how many of the things we do will never be missed. One example -- a fairly typical one -- is the old manual order-entry system used until the task was computerized five years ago, but still maintained "just in case." Another is the system of duplicate patient files that many hospitals maintain, one for billing, one for patient-care, each run on a different computer with a different program. Altogether, up to one-third of all clerical and control operations are likely to be found unneeded, because they either never served a purpose or because they have outlived it. And nothing is less productive than to make more efficient what should not be done at all.
The next question in respect to the two-thirds of operations that will be found to serve a need is: "What contribution to the business should each make? What purpose does it serve?" Managements usually think the answer to be obvious. But more often than not, no one has an answer; or the answer is patently wrong; or, worst of all, there is more than one.
"Why do we check all our sales peoples' expense accounts?" "To keep them honest, of course." But that is hardly a business objective. The right answer is: "To keep sales expenses under control." And this is best done -- and at a fraction of the cost -- by determining expense standards based, for instance, on a sales person's need to travel and on the number of nights spent away from home. All that is needed to arrive at these standards is for a small number of experienced sales people to keep a record of their actual expenses twice a year for one week.
The previous system in the company -- the system that thought its purpose was morality -- kept 11 clerks busy the year round. The new system employs not even one full-time person. And it further enabled the company -- a large national wholesaler of builders' supplies -- to cut its sales force to 158 people from 167, despite a steady growth in sales volume. Sales people have more time to sell, when they no longer misuse selling time filling out lengthy "swindle sheets."
Just as common as the wrong answer to the question of what purpose a given operation serves, are two or more answers. But a well-designed and cost-effective operation serves one purpose only. To combine two or more in one operation means inefficiencies and sky-high costs.
"We have two objectives in supplying our 2800 national distributors," said the builders' supply people in answer to the question of what their big logistics operation contributed. "We make sure that none of our distributors is ever out of stock. And we make sure that we don't pile up excess inventory." What was needed were two separate operations.
One would make sure that the distributors are always stocked adequately with the fast-moving standard items that together account for about half of the firm's dollar sales. This is being done by stocking the distributors with 15% to 20% more of these items than they will sell during the next three weeks. There is no central inventory of these items any more, and no inventory control. The stock level for every distributor is determined by systematic spot checks, taken every other week, of the actual retail sale of a 3% sample of distributors -- that is of 84 distributors throughout the country. This requires only seven or eight sales trainees, and has been found, incidentally, to be a most effective training tool.
A second operation then handles the 20% of "speciality" products -- mostly big-ticket items -- that together account for the other half of the firm's dollar sales (and for a substantially larger part of its profits). These are stocked in one central warehouse located at the hub of an air freight company, and shipped free of charge by overnight air delivery anyplace in the country within six hours of receipt of the order.
The old system cost almost 1% of the company's sales (and that in a business where a 6% return on sales is considered outstanding!). The new systems together cost less than one-third as much. And where the old system kept 53 people busy the two new systems together employ 20. Yet the new systems give both, better service and better inventory control.
The question of how to organize the restructured organization for maximum performance and minimum cost comes only at the very end. More computers to handle more data faster is rarely the right answer. To be sure, the end product will be in many cases expressed by a computer program. But the task is to define what information is needed rather than how to manipulate it.
This may mean -- as it did in one of my earlier examples -- switching from inside to outside data, in order to find out the actual retail sales of one's customers to the ultimate consumers. It may mean -- especially in operations aimed at controlling a process -- shifting from counting to statistics and sampling. Not only is sampling much cheaper than counting, it is far more reliable. Statistical analysis alone can provide the crucial information on which effective control rests: the difference between fluctuations within the permissible range of normal, and the "exception," that is the genuine malfunction, which calls for immediate remedy.
Cutting costs is only the beginning. If all that is being done is to cut costs without putting in adequate cost prevention, a recurrence of excess costs a few short years hence, can be guaranteed. For costs never drift down. Cost prevention requires steady work on productivity improvement of every operation, year in and year out -- with a 3% annual improvement a minimum goal. It requires that every operation and every activity be put, every third year or so, to the question: "Do we really need to do this or should it be abandoned?" It requires that new operations and activities -- and especially new staff operations -- be entered only if an old operation is abandoned or at least pruned back.
Each operation and activity should also be questioned -- again every three yearsas to the purpose it serves and the contribution it makes to the business. And each, finally, should be subjected to the question: What is the simplest way to achieve this purpose?
By now most of us have learned the hard way that dieting off fat is a good deal more difficult than not putting it on in the first place. Excess costs are excess fat. Cutting costs rarely gets much support from the work force itself; it means, after all, laying off people. Without active workforce participation, however, none of the measures needed for effective cost control are easy to implement. Indeed one reason why so many of the cost-cutting efforts of past years have failed to cut costs is that they were imposed from above on a work force that saw in them a threat to their own jobs and incomes. Cost prevention, however, can count on active, and indeed, enthusiastic work-force support. Employees know where the fat is. They also know that low, controlled costs mean better and more secure jobs.
A great deal more cost-cutting is still needed, especially in big organizations (and by no means only in American ones -- the big Japanese companies, e.g. the big banks, are far more over-staffed still). But cost-cutting should always be used as the first step towards building permanent cost prevention into the organization.