Late Night Discussions
Number 7

by Dr. Eliyahu M. Goldratt

I'm searching for a proper subject for tonight, but my mind is hooked on one sentence, 'Competition is fiercer than ever.... Competition is fiercer than ever...' Can I turn this obvious fact into a meaningful discussion?

"Jonah?" I ask, "In today's environment, where competition is fiercer than ever, do you think that we're right in concentrating on the production department?"

"No," he answers. "I think that industry should concentrate much more on engineering."

"Engineering?" I ask, quite surprised. I think about it for a while, and then say, "Do you mind if we examine it together? I can't see how engineering could possibly be more important than production."

"Let's examine it," Jonah rises to the challenge. "But Alex, do you mind starting the analysis at the beginning? We agree that competition for markets is fiercer than ever. Can you outline the factors that enable a company to better compete in the market?"

I don't have any problem answering, since we have already discussed this subject in depth. "A company can compete based on three different categories," I say confidently, "Its products, its prices and, not less importantly, its service."

"Agreed," says Jonah, and gestures for me to continue.

"Regarding the first category, the product," I start the comparison between production and engineering, "we have to be careful not to jump to the conclusion that it is totally in the hands of engineering. It's not only the capabilities of the product which count, the product quality is not less important. TQM has taught us that, contrary to the common excuses, product quality is solely in the hands of production."

Seeing Jonah's expression I ask, "Don't you agree that we should regard product engineers as part of production rather than engineering?"

When he doesn't answer I continue, "So, for the category of product, I would say, that production and engineering play equally important roles. As for the second category, I simply don't see how one can claim that, in general, the price is influenced more by engineering than by production."

Jonah's face shows his growing discomfort but it's probably just due to minor reservations since he doesn't interfere.

"That leaves us with the third category - service," I continue. "Here, as we've discussed so many times in the past, we must distinguish once again between two main issues, response time to the client request and due date performance - keeping the promises that we have given to the clients. As long as we are dealing with the more general case, in which no engineering changes are needed, I don't see how engineering is involved at all.

"All in all," I conclude, "I can't see how engineering could be more important than production, since in three important aspects, quality, response time and due date performance, engineering is not involved."

"Aren't they?" is the sarcastic reply.

This last remark of Jonah's punctures a hole in my euphoria. View it again, I say to myself. It is definitely the job of production to take care of quality, but does engineering put unnecessary obstacles in their way? Yes, of course. How many times have I seen production trying so hard to meet tolerances that shouldn't be there in the first place? This definitely doesn't help the product quality, and it jeopardizes both response time and due dates. But what can engineering do?

I allow my thoughts to wander. In order for production to produce, engineering must specify tolerances. True, at the time engineering has to specify these tolerances only some prototypes have been done; nobody knows what the tolerances really should be; it's more an educated guess than hard numbers. No wonder that engineering, in their eagerness to guarantee the product will function properly, puts very tight tolerances.

Maybe engineering should put two tolerances? A tight one that guarantees that the product will work, and a loose one that almost guarantees that the product won't function. Once production finds it difficult to meet the tight one, they can go back to engineering and the issue can be investigated. In such a case no one will have anything to protect.

This is a nice thought, and even amusing, but I can't believe that Jonah objected to my conclusion based on this relatively small issue.

"Jonah," I say, "if you still want to claim that engineering is more important than production in this fierce fight for the market, I'm afraid that you'll have to contribute more than just a sarcastic remark."

"Sorry," he apologizes. "You have done a very nice analysis but when you talk about fierce competition for the market, I think that you should address markets rather than clients. For example, take your third category, service - which you defined as including both response time and due date performance to a specific client. What will happen to the analysis if we define it more globally, as response time and due date performance to the market?"

"You have to be more specific," I smile. "I don't see the difference."

Jonah takes his time lighting his cigar but the extra time doesn't help, I still don't understand to what he is alluding.

"Alex," he starts at last, "when we talk about the time it takes to respond to the market needs, aren't we actually talking about the time that it takes a company to launch a better, new product?"

"Now I see," I answer, "why didn't you say that right from the start! Of course this response time is very important, especially in light of the fact that product life cycles are shrinking at breathtaking speed. In many industries we've already reached the point where the product life is significantly shorter than the time it takes engineering to develop the product. But Jonah, what do market due-dates have to do with engineering?"

"Alex, take for example a product that has an expected life of just one year before the more improved product will take over. What is the advantage of coming out with this product, let say, three months ahead of the competition?"

"In such an environment, three month is all it takes, the early bird can capture almost the entire market. I see, the due date to come out with a new product is actually set by the competitors. It's clear to me now that when competition has brought the product life to be roughly equal to the engineering time, these factors become dominant. It's no use having very good quality, price and customer service on an already obsolete product. Yes, it's apparent that the ball is now in engineering's court.

In a very patient voice he replies, "What have you learned in production? The major mistake was to ignore the combined effect of dependent events and statistical fluctuations. Where do we have more dependent events, for a product in production or for a project in engineering? Where do we have more statistical fluctuations, in production or in engineering, where surprises are the only thing that does not surprise anyone?"

"If you put it that way, definitely in engineering. But still, I can't envision a systematic way to manage engineering projects."

He sighs deeply, "If you are looking to copy, one to one, the methods from production, you're bound to fail. You have to start from first principles and develop the appropriate procedures. But don't fool yourself. As long as these two basic phenomena, dependent events and statistical fluctuations, exist, constraints which relate to both the project activity structure and to limited resources also exist.

"Just a moment, Jonah," I can't resist the opportunity to tease him, "before you drop the subject, are you telling me that we can defeat the rule that a project is never finished on time?"

"As far as I know this was never a rule of nature, it is just a comfortable excuse," he smiles back.

This "Late Night Discussion" is Copyright © 1991 Dr. Eliyahu M. Goldratt