Yes, recently I was invited to take yet another "Training Course," following in the path of great successes such as Taguchi Methods, Fuzzy Logic, and Time Management. But why would this course on basic accounting and economics be called "Apples and Oranges"? Because, the course's leader explained, if one person has apples and another person has oranges, the only way to quantify the sum total of the two fruits is their value in dollars. So we quickly dropped the old aphorism that "you can't compare apples and oranges," and started talking about dollars.
We soon agreed that it was a good idea for our company to be profitable, and that all of us employees ought to be aware of how to improve profitability. But, how is the profit to be measured—how do we do the accounting? We started out with a little game from Learning Methods International, Box 50, S-230-42 Tygelsjo, Sweden. The purpose of the game is to teach you which expenditures of a manufacturing plant are used to compute the profits, and how the costs and expenses and assets are added and subtracted and divided to compute the Return On Investment. Michael, the leader, gave us an explanation of how the "game" is supposed to run—a discussion of how this little hypothetical "factory" buys raw materials and adds labor to accomplish the manufacturing of Widgets. Then the Widgets were put into "Finished Goods" to await customer orders. Over the course of a year, this factory was able to show a good operating profit. But after expenses for interest and taxes, the after-tax profit was only $2 million—not a very good profit on $80 million of sales.
Michael suggested that we go back and recompute the profits if we went back and followed a different game plan: We should pay off $20 million of our liabilities (bank loans) with some of our cash. Sure enough, at the end of the year, the net profits looked considerably better, about $3 million. Further, he suggested that we should not keep a lot of our finished products in Finished Goods (or, "Box Stock," as we call it at National), but we might plan to just build our Widgets to meet customer orders. We could save a lot of money by not having a 3-month stock of expensive assembled parts waiting around for customer orders. FURTHER, we could save additional money by not buying our raw materials from just any cheap suppliers, but rather by choosing a "small number of very reliable suppliers" that we could trust to ship us our raw materials on a JIT (Just-In-Time) basis. This saves more money by not having raw materials sitting in our storehouse for a long time. Altogether, after we ran the numbers through the exercise, our new lean, mean company had made $6 million of net profit. Wasn't that a lot better, asked Michael?
I spoke up, "HELL, NO! The only reason we were able to save $4 million in interest was because this whole scenario was set up badly. In the base example, the company treasurer sat on an average of $51 million of cash, and never even put it in the bank to earn daily interest. The only way we could show an improvement of $4 million on interest was to ignore the fact that the treasurer had wasted $8 or $10 million in interest he could have earned. Even in the second scenario, there was $3 or $4 million left lying on the table due to the treasurer's ignorance and/or incompetence.
"It's absurd to argue that one plan shows superiority if it's only compared to a Straw Man." (As you can see, I am not a big fan of Straw Man comparisons....) So the lecturer immediately spotted me as a troublemaker. "Well, yes, you are right, but we just have to follow these examples and procedures for accounting, which are, of course, simplified."
Next I complained that if you order your raw materials from "a limited number of highly reliable suppliers" for the purpose of getting "JIT delivery," you should honestly expect to pay more for that. I've heard that half the trucks in Tokyo are driving around in circles, half-empty, wasting time so they can show up at JUST the right minute, neither late nor early.
This might make your assembly operations simpler, but that JIT service is NOT free, and not necessarily cheap. Somewhere, somehow, you will pay for it. Further, one traffic jam, one flat tire, and your production line goes down. "Yes, but we are just following our guidelines, which are of course simplified..."
Later, I pointed out, after the game had been played for three years of financial analysis, "Isn't it kind of stupid to make only $80 million a year of these Widgets? What the heck is going on? In most cases, if you spent another $1 million to hire some better salesmen, you could sell a lot MORE than $80 million, and the increased profits would more than pay for the cost of sales." NOW, I must admit, in a hypothetical example, maybe we can't be more specific on what a Widget really is. So if a "Widget" is a kind of commodity such that, no matter how many you have, no matter how smart the salesmen are, they could not sell more than the customer was planning to buy—then we should not insist on hiring better salesmen. But otherwise, we would be remiss if we didn't study that possibility. The Instructor agreed that, in some cases, expanding sales is indeed the right thing to do, but there was no room to show that in this simplified example.
After Michael begged me not to be such a disruptive influence during the class, I met with him during lunch hour. I pointed out that in his "simplified" example, he claimed that he was able to sell the same $80 million of Widgets by manufacturing them on a Just-In-Time basis. But at NSC, we often found customers who WANTED to buy a product from stock RIGHT AWAY, but if we had nothing in stock, they would have to buy elsewhere.
So IF you're in a business in which you're SURE that nobody will buy MORE parts just because they are on the shelf, and you're SURE nobody will buy LESS just because the parts are NOT on the shelf, well, go ahead and do your JIT manufacturing. But I warned him that this OVERSIMPLIFICATION was NOT necessarily good for business, NOT good for profits, and NOT valid for improving ROI. He conceded that my arguments might have some truth—but there was no room for them in this simplified analysis.
One of the other students said, "Well, Bob, you may not like the simplified stuff in this course, but it sure got you thinking." I replied that I had been thinking about this very seriously for several years, and I already figured out some of the weaknesses of JIT manufacturing. Hearing this guy just preaching oversimplified economics got me so ticked off that I was going to do something about correcting this problem. And even though your company may not make the same kind of products as mine, you probably face the same kinds of issues.
Where I work, I'm often concerned about how we can avoid a complaint from a prospective customer: "How can I design in your parts right away if your samples are only available on a three-week delivery from Malacca?" So I worked pretty hard with our planners to maximize our ability to ship evaluation samples overnight, or in 1 or 2 days. But I have to fight with some guys who think that keeping our samples in the low-rent district of Timbuktu is a great way to save money. I certainly don't need a training course to simplify things to the point where people think that the cheapest costs are always the right solution.
Furthermore, I do get calls like—"After I get your new parts designed in, how can I go into pilot production if the availability is 6 weeks?" I point out to them that while we may not necessarily have these parts in stock, we have distributors who do a pretty good job of keeping at least small or medium quantities in stock, so you can order them and get prompt delivery. We want to keep a lean (but not zero) inventory.
Now, if a guy complains "I ordered the ICs for my first production run; then the product got HOT and I wanted to double my production run—but the availability of parts was not good. What do I do now?" I've been arguing for years that we should make sure our sole-sourced parts, our proprietary ICs, and, for that matter, any of our products that have a decent volume and good profitability are kept in stock in some reasonable quantity. The alternative is to tell your customers, "We can't help you; you will have to design in our competitor's parts." So when I heard this nincompoop at the training session arguing that we should show a better ROI by only selling Widgets to people who can place orders well in advance of when they needed them, I decided to start chomping my teeth into peoples' ankles to straighten out this oversimplified economic foolishness.
One copy of my first draft went right to our Chief Executive Officer, Gil Amelio, and I have also brought it to the attention of lots of other people in sales, marketing, and planning areas. These guys have given me much support and encouragement. Maybe Michael doesn't want any loose cannons in his lecture room, but in the real world, if somebody else doesn't complain about stupid, oversimplified strategies, then it becomes my job. Go ahead, tell me that I'm a loose cannon. We all already knew that. Even a loose cannon can do some good if it sweeps away cobwebs. Maybe I can't tell the difference between apples and oranges, but I can tell the difference between bad economic theory, OVERSIMPLIFICATION, and the real world.
Now, speaking of oversimplification, when I propose keeping these new, proprietary or sole-sourced parts in stock, I shouldn't imply that it's easy to do. Murphy's Law says that as soon as I see a part selling with good volume, and I put a fat amount of them in stock, then people will stop buying them. If I put the expensive one in stock, people will want to buy the cheap ones (and vice-versa); if I stock the mini-DIPs, engineers will specify the surface-mount versions. So if I indicated that keeping the right balance of parts in stock would be easy, I must apologize for my own oversimplification.
We have to try to anticipate our customers' ordering patterns. We have to try to keep a reasonable bank of good, tested dice, so no matter which packages the customers ask for, we can assemble the right ones quickly. To do this promptly may require an agile, fast-response assembly group. This may cost a little more, but the advantages may well be worth it in terms of being able to respond and fulfill customer needs.
Also, that doesn't mean that we just have to keep lots of parts in our store rooms. If one distributor is keeping one part in stock, and another is stocking another part, then we may not have to keep many of those parts around at all. But we have to keep a few around for quick sampling.
I talked with some of the guys who designed the original three-day comprehensive lecture about reality and economics. They were surprised to see how unrealistic and badly oversimplified the economic studies became when they were pared down to a one-day course. They agreed it was probably time to see what we could do to restore some reality to that little course. I sent in five scribbled pages of my suggestions on how to restore some of the redeeming factors. I also invited the accountants who really know what's important in fiscal matters to add comments about things I didn't notice, or am not very knowledgeable about. Maybe next year, the "Apples and Oranges" course will not only be "training," but will really educate us on how to make our company more profitable! The best way for all you managers out there (and other workers, too) to please your customers may involve thinking—NOT just referring to some old obsolete formula or cookbook.
Comments invited! / RAP
Robert A. Pease / Engineer
P. S. Mark Levi, who was involved in the design of the original Apples and Oranges course, and who helped me a lot in getting this column into good shape, suggested I read a book: The Goal, by Eliyahu M. Goldratt and Jeff Cox. This is written in the form of a novel about a guy who is trying to run a manufacturing plant in the midst of all sorts of trouble. His customers are screaming about late delivery, his workers are setting contradictory rules, his most efficient machines are being misused, and his boss is threatening him.
Meanwhile the bean counters tell him everything he does to improve the situation is wrong. A very stressful job—trying to bring in a few more dollars to show his boss why the plant should not be shut down. This is a COMPLETE antithesis of Apples and Oranges, that Pollyanna-ish waste of time. However, Mr. Goldratt doesn't gratuitously overcomplicate things—he sets reasonable examples and covers one or two problems at a time. Good thinking, good writing. The book was published in 1986 by North River Press, Box 567, Great Barrington, MA 02130.
You can order by calling (413) 528-0034 or (800) 486-2665. Price is about $19.95. Bravo to a good story-teller, explaining how the real world is horribly different from those orderly, sterile optimization situations you learn about in school. —RAP
Originally published in ELECTRONIC DESIGN, May 2, 1994
RAP's 2000 comments: No additional comments. I said it all the first time. Many readers agreed.—rap