Quality has become a business fad of the 1980s. Books, television, magazines, and newspapers regularly feature people who've become gurus of quality. Seminars, tapes, pamphlets, journals, wall posters, and bumper stickers trumpet the word: Quality! Quality! With a capital Q, and that rhymes with you, and that. . . .

Wait a minute. Aren't we getting just a little carried away? Sure, quality's important. But right now, you've got orders to fill, payroll to meet, cash to raise. Right now, you've got to survive. When things calm down, maybe then you can worry about quality.

I recently came across an old memo that reminded me how closely quality can be tied to business survival. When I wrote it, I was vice-president of finance for a company I'll call Fubar. This company had invented a product everybody wanted -- immediately. And Fubar had to satisfy those demands. Without immediate shipments, it wouldn't have the cash to survive.

Our engineering and manufacturing departments worked night and day to ship on schedule. But in their urgency, they lost control of the quality of our purchased parts. Soon after we had begun to ship, therefore, the low-quality parts created a cash-flow nightmare. I wrote the following memo to the president to summarize the causes of this nightmare. Perhaps it will help you recognize similar problems that may be developing in your own company.

We have a cash-flow problem caused by poor-quality parts. These consume Fubar cash in the following ways:

1. They frequently delay the shipment of our products.

Because we were so short to cash, we had negotiated COD payment terms with our larger customers. But it was common for a $500,000 shipment -- representing desperately needed cash -- to sit in final assembly waiting for hundreds of dollars' worth of parts to be reworked to fit.

2. When parts are rejected, the inventory of assemblies requiring these parts builds up in work in process, increasing our overall inventory.

Manufacturing had two reasons for building assemblies before having a complete inventory of usable parts on hand. First, we needed shipments so badly that our manufacturing people wanted to be able to ship as soon as possible after they received the missing parts. Second, since we were so desperate for working parts, manufacturing often ignored normal inspection procedures. If there were any chance that a part would work, it passed through inspection to work in process. Often, assemblies were nearly completed before manufacturing discovered an unusable part that would stop production.

Later this problem snowballed when someone realized that if we stripped part A from one batch of assemblies and part B from another batch, we could install the two parts in still another batch and actually get a product out the door. Occasionally, therefore, we paid overtime the first day to install a part, the second day to remove the part, and the third day to install it in a different assembly for shipping. And most of the scavenged assemblies were eventually scrapped.

3. The purchasing department orders more parts than we need for production because:

a) It's faster to order new parts than to fix the old.

Many of the problems were caused by Fubar. Our engineers occasionally gave vendors incorrect dimensions, or specified parts informally, using hand-drawn sketches and best-guess dimensions. If there'd been time, manufacturing could have reworked those parts to fit. But since there was no time, it set them aside "temporarily" and ordered more.

b) Purchasing expects a high fallout and wants to make sure it has enough usable parts to meet our production schedule.

During this period, purchasing agents were fired for not having parts on hand when needed. Rarely was anyone disciplined for ordering too many.

4. In the past, we have been slow to return rejected parts to the vendor. We've thus carried a double inventory.

I understated the case here. We had many months of rejected parts on hand, for several reasons. First, manufacturing had to decide whether the problem was our fault or the vendor's. If it was the vendor's, manufacturing had to document that fact. Second, dimensions changed frequently. Before manufacturing could determine fault, it had to determine the dimension used on the day the part was ordered. This was often an impossible task. And third, no one had time to mess with a growing mountain of rejected parts.

5. The more rejected parts we have, and the longer we hold them, the greater the number of parts we'll damage ourselves. And it's tough to return these doubly damaged parts to vendors for credit.

It was easy enough to see the problem from the vendor's viewpoint. "Before we drove a forklift over these parts, Mr. Vendor, the paint contained hairline cracks, making the parts unusable. Like this part here, with the tread marks on it. . . ."

6. The more rejected parts we have, the more reasons we find to scrap or rework parts at Fubar expense.

The primary responsibility of Fubar's manufacturing manager was to get product out. He had a pile of rejected parts in the factory that could soon grow to the size of a small continent. If some of those parts could be reworked, and he needed it done immediately, he'd gladly pay. As for the other parts, there was no time to assign blame, then negotiate with vendors. Besides, even if the vendors were at fault, he'd bankrupt the smaller ones if he didn't pay for that mountain of trash -- and he needed those vendors.

7. At times, we accept bad parts, rework them, then adapt them to other uses. For example, the A103 is a drive pulley used on most of our products. When the sonic weld on this pulley fails, we use it as a spacer on the Model 12. So we pay for a costly process that doesn't work, and rather than getting our money back, we use the expensive part as a cheap one.

This was the only instance I found where an expensive part was substituted for a cheap one. But it had become such an accepted practice in manufacturing that no one could understand my shock when I discovered it. To me, this practice represented the depths to which the quality problem had taken us.

8. Rejected parts place a significant administrative burden on the organization. Purchasing, accounting, quality control, manufacturing, shipping, receiving, and other departments spend frust rating and expensive hours dealing with problems created by low-quality parts.

In accounts payable, for instance, one person worked full-time with our two worst vendors, processing debit and credit memos, return-to-vendor paperwork, corrected invoices, and marked-up receiving reports. The middle manager who was responsible for processing rejected parts worked 12-hour days for months on end making no apparent progress. When that manager quit, no one would take his job.

9. To meet production schedules, we must accept parts that are not to specification. This causes the following additional costs:

a) We spend many hours in reworked and slowed production.

To deal with the quality problems "efficiently," each assembly worker was issued metal files and cans of spray paint. If parts didn't fit, workers were expected to file them down, then cover up the damage with paint.

b) We spend many additional dollars in secondary processing -- painting, for example -- to cover up the problems. This causes more quality problems because paint scratches more easily than the original surface.

Even if we ignored the cost of the additional rework and secondary processing, the effect of this work on our cash was substantial. At times, Fubar inventory was spread among 10 vendors within a 50-mile radius, where the parts were ground, milled, tumbled, welded, puttied, and painted. Even though one employee tracked this work full-time, I was never sure where my inventory was or what condition it was in.

c) Possible deterioration of Fubar product quality.

I remember rewriting this item several times, making it more wishy-washy each time. Our president talked a lot about Fubar's high-quality product and wouldn't abide suggestions to the contrary. When I mentioned quality to a grizzled manufacturing foreman one day, the man swore and said, "I may look like Rumpelstiltskin, but I can't spin crap into gold."

We have so many bad parts to process right now that it's easy to lose sight of our goal. Our goal should not be to gain greater efficiency in processing poor-quality parts. Our goal should be to reduce, then eliminate, the incidence of these parts in the first place.

Not long ago, Philip B. Crosby wrote a fascinating book called Quality Is Free. The title is wrong, however. Based on my experience, quality isn't just free. It pays.