"There you are," he says.
We read it.
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
139
"Okay, so you only got nineteen pieces done in the first
hour," I say .
"Well, it took us a little longer to get organized, and one guy
was late coming back from lunch," says Pete. "But at one o'clock we had a materials handler take the nineteen over to the robot so
it could get started."
"Then from one to two, you still missed the quota by four
pieces," says Bob.
"Yeah, but so what?" says Pete. "Look what happened from
two o'clock to three: we beat the quota by three pieces. Then
when I saw we were still behind, I went around and told every-
one how important it was for us to get those hundred pieces done
by the end of the shift."
"So everyone went a little faster," I say.
"That's right," says Pete. "And we made up for the slow
start."
"Yeah, thirty-two pieces in the last hour," says Bob. "So what do you say, Al?"
"Let's go see what's happening with the robot," I say.
At five minutes past five o'clock, the robot is still turning out
welded sub-assemblies. Donovan is pacing. Fred walks up.
"Is that truck going to wait?" asks Bob.
"I asked the driver, and he says he can't. He's got other stops
to make and if he waits for us, he'll be late all night," says Fred.
Bob turns to the machine. "Well, what the heck is wrong with
this stupid robot? It's got all the parts it needs."
I tap him on the shoulder.
"Here," I say. "Look at this."
I show him the sheet of paper on which Fred has been re-
cording the output of the robot. From my shirt pocket, I take out
Pete's log and fold the bottom of it so we can put the two pieces of
paper together.
Combined, the two of them look like this:
I tell him, "You see, the first hour Pete's people did nineteen
pieces. The robot was capable of doing twenty-five, but Pete deliv-
ered less than that, so nineteen became the robot's true capacity
for that hour."
"Same with the second hour," says Fred. "Pete delivered
twenty-one, the robot could only do twenty-one."
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
140
Output = 90 pcs.
"Every time Pete's area got behind, it was passed on to the
robot," I say. "But when Pete delivered 28 pieces, the robot could still only do twenty-five. That meant that when the final delivery
of thirty-two pieces arrived at four o'clock, the robot still had
three pieces to work on from the last batch. So it couldn't start on
the final batch right away."
"Okay, I see now," says Bob.
Fred says, "You know, the most Pete was ever behind was ten
pieces. Kind of funny how that's exactly the number of pieces we
ended up short."
"That's the effect of the mathematical principle I was trying
to explain this morning," I say. "The maximum deviation of a
preceding operation will become the starting point of a subse-
quent operation."
Bob reaches for his wallet.
"Well, I guess I owe you ten bucks," he says to me.
"Tell you what," I say. "Instead of paying me, why don't you give the money to Pete so he can spring for a round of coffee or
something for the people in his department—just a little way to
say thanks for the extra effort this afternoon."
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
141
"Yeah, right, that's a good idea," says Bob. "Listen, sorry we couldn't ship today. Hope it doesn't get us in trouble."
"We can't worry about it now," I tell him. "The gain we
made today is that we learned something. But I'll tell you one
thing: we've got to take a close look at our incentives here."
"How come?" asks Bob.
"Don't you see? It didn't matter that Pete got his hundred
pieces done, because we still couldn't ship," I say. "But Pete and his people thought they were heroes. Ordinarily, we might have
thought the same thing. That isn't right."
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
142
18
When I get home that evening, both of the kids greet me at
the door. My mother is in the background, with steam pouring
out of the kitchen. I presume it has something to do with dinner
and that she has everything under control. In front of me,
Sharon's face is beaming up at me.
"Guess what!" she says.
"I give up," I say.
"Mommy called on the phone," Sharon says.
"She did!" I say.
I glance up at my mother. She shakes her head.
"Davey answered the phone," she says. "I didn't talk to her."
I look down at Sharon. "So what did Mommy say?"
"She said she loved Davey and me," says Sharon.
"And she said she would be away for a while," adds Davey.
"But that we shouldn't worry about her."
"Did she say when she would be coming back?" I ask.
"I asked her that," says Davey. "But she said she couldn't say right now."
"Did you get a phone number so I can call her back?" I ask
him.
He looks down at the floor.
"David! You were supposed to ask her for the number if she
called!"
He mumbles, "I did, but . . . she didn't want to give it to
me."
"Oh," I say.
"Sorry, Dad."
"It's okay, Dave. Thanks for trying."
"Why don't we all sit down to dinner," my mother says
cheerily.
This time the meal is not silent. My mother talks, and she
does her best to cheer us up. She tells us stories about the Depres-
sion and how lucky we are to have food to eat.
Tuesday morning is a little bit more normal. Joining efforts,
my mother and I manage to get the kids to school and me to
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
143
work on time. By 8:30, Bob, Stacey, Lou, and Ralph are in my
office, and we're talking about what happened yesterday. Today,
I find them much more attentive. Maybe it's because they've seen
the proof of the idea take place on their own turf, so to speak.
"This combination of dependency and fluctuations is what
we're up against every day," I tell them. "I think it explains why we have so many late orders."
Lou and Ralph are examining the two charts we made yes-
terday. "What would have happened if the second operation
hadn't been a robot, if it had been some kind of job with people?"
asks Lou.
"We would have had another set of statistical fluctuations to
complicate things," I say. "Don't forget we only had two opera-
tions here. You can imagine what happens when we've got de-
pendency running through ten or fifteen operations, each with
its own set of fluctuations, just to make one part. And some of our
products involve hundreds of part
s."
Stacey is troubled. She asks, "Then how can we ever control
what's going on out there?"
I say, "That's the billion-dollar question: how can we control
the fifty-thousand or—who knows?—maybe it's fifty-million vari-
ables which exist in this plant?"
"We'd have to buy a new super computer just to keep track
of all of them," says Ralph.
I say, "A new computer wouldn't save us. Data management
alone isn't going to give us more control."
"What about longer lead times?" asks Bob.
"Oh, you really think longer lead time would have guaran-
teed our ability to ship that order to Hilton Smyth's plant?" I ask
him. "How long had we already known about that order before
yesterday, Bob?"
Bob wiggles back and forth. "Hey, all I'm saying is that we'd
have some slop in there to make up for the delays."
Then Stacey says, "Longer lead times increase inventory,
Bob. And that isn't the goal."
"Okay, I know that," Bob is saying. "I'm not fighting you.
The only reason I mention the lead times is I want to know what
we do about all this."
Everybody turns to me.
I say, "This much is clear to me. We have to change the way
we think about production capacity. We cannot measure the ca-
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
144
pacity of a resource in isolation. Its true productive capacity de-
pends upon where it is in the plant. And trying to level capacity
with demand to minimize expenses has really screwed us up. We
shouldn't be trying to do that at all."
"But that's what everybody else does," says Bob.
"Yes, everybody does. Or claims to. As we now can see, it's a
stupid thing to try," I say.
"So how do other manufacturers survive?" asks Lou.
I tell him I was wondering that myself. What I suspect is that
as a plant comes close to being balanced through the efforts of
engineers and managers doing the wrong things, events head
toward a crisis and the plant is very quickly un balanced by shifting workers or by overtime or by calling back some people from
layoff. The survival incentive overrides false beliefs.
"Okay, but again, what are we going to do?" asks Bob. "We
can't hire without division approval. And we've even got a policy
against overtime."
"Maybe it's time to call Jonah again," says Stacey.
And I say, "I think maybe you're right."
It takes Fran half an hour to locate the area of the world
where Jonah happens to be today, and another hour passes be-
fore Jonah can get to the phone to talk to us. As soon as he's on
the line, I have another secretary round up the staff again and
corral them in my office so we can hear him on a speaker phone.
While they're coming in, I tell Jonah about the hike with Herbie
where I discovered the meaning of what he was telling me, and
what we've learned about the effects of the two phenomena in the
plant.
"What we know now," I tell him, "is that we shouldn't be
looking at each local area and trying to trim it. We should be
trying to optimize the whole system. Some resources have to have
more capacity than others. The ones at the end of the line should
have more than the ones at the beginning—sometimes a lot
more. Am I right?"
"You're on the money," says Jonah.
"Good. Glad to hear we're getting somewhere," I say. "Only
the reason I called is, we need to know where to go from here."
He says, "What you have to do next, Alex, is distinguish
between two types of resources in your plant. One type is what I
call a bottleneck resource. The other is, very simply, a non-bottle-
neck resource."
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
145
I whisper to everybody to start taking some notes on this.
"A bottleneck," Jonah continues, "is any resource whose ca-
pacity is equal to or less than the demand placed upon it. And a
non-bottleneck is any resource whose capacity is greater than the
demand placed on it. Got that?"
"Right," I tell him.
"Once you have recognized these two types of resources,"
says Jonah, "you will begin to see vast implications."
"But, Jonah, where does market demand come in?" Stacey
asks. "There has to be some relationship between demand and
capacity."
He says, "Yes, but as you already know, you should not bal-
ance capacity with demand. What you need to do instead is bal-
ance the flow of product through the plant with demand from the
market. This, in fact, is the first of nine rules that express the
relationships between bottlenecks and non-bottlenecks and how
you should manage your plant. So let me repeat it for you: Balance flow, not capacity."
Stacey is still puzzled. She says, "I'm not sure I understand.
Where do the bottlenecks and non-bottlenecks come into the pic-
ture?"
Jonah says, "Let me ask you: which of the two types of re-
sources determines the effective capacity of the plant?"
"It would have to be the bottleneck," she says.
I say, "That's right. It's like the kid on that hike last weekend
—Herbie. He had the least capacity and he was the one who
actually determined how fast the troop as a whole could move."
"So where should you balance the floor?" asks Jonah.
"Oh, I see," says Stacey. "The idea is to make the flow
through the bottleneck equal to demand from the market."
"Basically, yes, you've got it," says Jonah. "Actually, the flow should be a tiny bit less than the demand."
" "How come?" asks Lou.
"Because if you keep it equal to demand and the market
demand goes down, you'll lose money," says Jonah. "But that's a
fine point. Speaking fundamentally, the bottleneck flow should
be on a par with demand."
Bob Donovan is now making various noises, trying to get
into the conversation.
"Excuse me, but I thought bottlenecks were bad," says Bob.
"They ought to be eliminated where possible, right?"
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
146
"No, bottlenecks are not necessarily bad—or good," says Jo-
nah, "they are simply a reality. What I am suggesting is that
where they exist, you must then use them to control the flow
through the system and into the market."
That makes sense to me as I'm listening, because I'm remem-
bering how I used Herbie to control the troop during the hike.
"Now I have to run," says Jonah, "because you caught me
during a ten-minute break in a presentation."
I jump in. "Jonah, before you go—!"
"Yes?"
"What's our next step?"
He says, "Well, first of all, does your plant have any bottle-
necks?"
"We don't know," I tell him.
"Then that's your next step," he says. "You have to find this out, because it makes an enormous difference
in how you manage
your resources."
"How do we find the bottlenecks?" says Stacey.
"It's very simple, but it would take a few minutes to explain.
Look, try to figure that out for yourselves," says Jonah. "It's really easy to do if you think about it first."
I say, "Okay, but. . . ."
"Good-bye for now," he says. "Call me when you know if you
have a bottleneck."
The speaker phone issues a click, followed by a fuzzy hum.
"Well . . . what now?" asks Lou.
"I guess we look at all our resources," I say, "and compare them against market demand. If we find one in which demand is
greater than capacity, then we'll know we've got a bottleneck."
"What happens if we find one?" asks Stacey.
"I guess the best thing to do would be what I did to the scout
troop," I say. "We adjust capacity so the bottleneck is at the front of production."
"My question," Lou says, "is what happens if our resource
with the least capacity in fact has a capacity greater than what
market demand calls for?"
"Then I guess we'd have something like a bottle without a
neck," I say.
"But there would still be limits," says Stacey. "The bottle would still have walls. But they'd be greater than the market demand."
E.M. Goldratt
The Goal: A Process of Ongoing Improvement
Captured by Plamen T.
147
"And if that's the case?" asks Lou.
"I don't know," I tell him. "I guess the first thing to do is find out if we've got a bottleneck."
"So we go look for Herbie," says Ralph. "If he's out there."
"Yeah, quick, before we talk ourselves to death," says Bob.
I walk into the conference room a few days later and there's
paper everywhere. The main table is covered with computer
print-outs and binders. Over in the corner, a data terminal has
been installed; next to it, a printer is churning out even more
paper. The wastebaskets are full. So are all the ashtrays. The litter
of white styrofoam coffee cups, empty sugar packets and creamer
containers, napkins, candy bar and cracker wrappers, and so on
is scattered about. What has happened is the place has been
turned into our headquarters in the search for Herbie. We have
not found him yet. And we're getting tired.
Sitting at the far end of the main table is Ralph Nakamura.
He and his data processing people, and the system data base they
manage, are essential to the search.
Ralph does not look happy as I come in. He's running his
The Goal: A Process of Ongoing Improvement, Third Revised Edition Page 19
