Notes on "The Goal" by Goldratt

When people fill up their schedules, that typically uses up additional bottlenecked resources like time or energy.

1 Like

29

alex is sleeping with julie but he’s having trouble keeping his mind off the plant.

the plant has got half a dozen new orders. they timed release of new materials for the bottleneck and efficiencies dropped for a while until they consumed excess inventories, but then they increased.

they implemented smaller batch sizes two weeks ago. before this non-bottlenecks were sometimes idle cuz they were waiting for a previous station to finish a batch. but since the batches are smaller the stations have less idle time between batches. so the idle time is more spread out and there is less idle time so efficiencies are higher.

inventories are lower too since the only work in progress is for current demand.

but cost of parts looks like it has gone up because of more setups cuz of smaller batch sizes, which doesn’t matter for the bottom line but looks bad politically. same setup time for fewer parts leads to more labour costs per part.

since they’ve reduced inventory and increased sales that labour cost is spread over more product so the cost of producing the product has decreased not increased.

measurement of labour costs assumes all the labour is fully occupied so more setups = more labour = more cost, but that’s wrong.

alex goes to work and meets lou. lou sez he can use cost of parts for the past two months instead of the past 12 months which is more representative and reduces the cost of parts problem. but that’s not standard accounting practice and ethan frost might not like it if he finds out.

jons calls and sez that bucky burnside wants 1000 model 12s in two weeks, which would be $1 million in sales. a competitor hadn’t filled it for 5 months and burnside had heard we’re turning orders around faster. filling the order could lead to them becoming burnside’s preferred supplier.

alex sez he will try to work out if they can do it. he has a meeting with bob, stacey and ralph. they think they could do it if they don’t work on anything else, but that would piss off other customers.

ralph suggests shipping some orders on time instead of early which would free up some bottleneck time and not create problems.

alex thinks that if they halve batch sizes again they might get higher throughput and be able to fill the order.

stacey sez a supplier isn’t going to be able to fill the order for 4-6 weeks for a part they can’t make in the plant. alex asks if they can stagger the shipments and use air mail to get the parts faster.

Alex tells jons he can do 250 units per week for four weeks, starting from two weeks after they get the order.

burnside accepts and sez he likes getting the model 12s in batches of 250 more than they like getting them all at once.

1 Like

30

they have a staff meeting.

they delivered the first shipment of the burnside order.

the smaller batches have increased the flow.

they’re up to 17% not just 15%

inventories are 40% of what they were three months ago and throughput has doubled.

bill peach congratulates alex on the plant performance and invites him to a review of the plant.

next week alex is offsite at a meeting.

hilton smyth has come to the plant to do a video of the robots in his capacity as productivity manager guy.

he was going to film in front of a robot for stacking stuff but it wasn’t running cuz it had no inventory and none on the way.

then smyth started asking questions about batch sizes.

two days later an audit team comes from hq and finds out that they’ve been calculating costs over the past two months rather than over the past year. when they recalculate it looks like costs have gone up.

johnny jons comes to the plant with bucky burnside. the last shipment was on time and had no quality problems. burnside came to congratulate the plant staff.

jons said burnside is going to order 10000 model 12 units per year.

alex goes to see julie the day before the plant review. julie sez they’ve drifted apart but she wants more of him in her life not less. she hasn’t wanted to go back to bearington because he’s paying more attention to her now. they agree to talk to one another for a couple of hours each night about their problems.

31

alex goes to his review: it’s being run by hilton smyth and neil cravitz not bill peach.

alex explains that the goal of the plant is to make money not cut costs, then explains bottlenecks, inventory etc.

smyth sez this is common sense. alex sez it isn’t cuz he’s balancing flow with demand not the capacity. also the utilisation level of a worker isn’t determined by his individual potential but by bottlenecks.

alex sez an hour saved at a bottleneck is an hour saved for the system. an hour saved anywhere else is useless. hilton smyth insists that an hour saved anywhere saves time and that bottlenecks only limit throughput temporarily, but they don’t have much impact on inventory.

cravitz asks how they can measure the performance of the plant if not by efficiencies. alex sez it is measured by the bottom line.

hilton smyth sez the plant can’t be profitable for long if cost of products increase and profits have to decrease.

alex goes to bill peach directly to say smyth is going to makes a negative recommendation and explains why smyth is wrong.

peach calls in smyth, cravitz and johnny jons, and asks smyth for his recommendation.

smyth sez alex’s plant should be fixed so that cost per product decreases.

ethan frost sez the plant made a profit.

jons sez that alex’s plant delivers stuff in a short time and help blast the market.

smyth sez this is temporary and they should expect large losses in the future.

bill peach sez the plant won’t be closed cuz it’s making a profit.

smyth protests.

bill peach sez that he has met with granby and is being promoted and can pick his successor. alex will take over from him as division manager.

alex goes back to the plant and calls jonah to ask for help.

jonah sez alex should rely more on himself and should decide what he wants to know before calling again.

alex sez he wants to run an efficient division.

jonah sez previously alex wanted to run an efficient plant, now he wants to run an efficient division and it won’t stop there.

alex sez he wants to learn how to manage any size of organisation and his own life.

jonah sez he should learn what techniques are available for effective management.

32

julie wants to toast alex’s promotion. alex feels his family paid too high a price for the promotion. julie sez the plant had problems and he had to solve them.

alex tries to give jonah credit, but julie points out alex came up with solutions himself. alex sez jonah gave him solutions in the form of pointed questions.

julie asks why alex didn’t think up the questions himself. alex sez jonah’s ideas seemed like common sense, but you only recognise ideas as common sense if they match your intuition.

alex sez he tried to explain jonah’s ideas but jonah’s practice of asking questions worked better.

if you try to explain stuff the person either understands what you’re saying or he doesn’t. if he doesn’t understand he’ll ignore you. if he does understand then he’ll think your point is below the belt.

but when alex tried to ask questions they were perceived as patronising.

jonah’s method of asking questions depends on him asking the right questions but also knowing what questions the other person will have.

alex wonders when jonah had time to learn so much about the shop floor when he was an academic.

33

alex wants lou to come with him to his new job, but thinks he won’t want to cuz he’s close to retirement and active in his community.

alex asks how much profit the plant will make and lou sez 20%.

lou sez he found a distortion in how they calculate profit that’s larger than the problem they had with product costs.

lou points out that inventory is a liability but it’s reported as an asset. wip inventory went down by 50% and finished goods inventory by 25% but this was recorded as a loss of assets and was only partly offset by the fact that they didn’t buy as much new materials.

when the inventory is put down a liability the plant made over 20% profit for the past 3 months.

everyone counts inventory as an asset, including the tax man.

lou sez he explained this to ethan frost, which is why frost supported him at the plant review.

lou sez he’s been thinking about the fact that he’s got old ideas and is close to retirement. he sez he’s been using bad procedures without understanding that he was doing harm. he thinks financial measurements should give them knowledge of to what extent they’re succeeding at making money and should make the parts of the business work together to serve the business as a whole. but current procedures don’t do that.

lou sez he wants to some with alex and develop a new financial measurement system that serves those objectives.

alex asks bob Donovan if he wants to be head of production for the whole division and he declines.

bob sez he learned so much from the changes introduced by the burnside order that he wants to stay at the plant and develop those ideas further.

instead of just saying yes or not to burnside they came up with a new offer based on what they could do and the client preferred it. bob thinks the lead time should be quoted case by case depending on the bottlenecks. and they shouldn’t always supply finished products in one batch. he’s been working on it with stacey and ralph.

but he can’t do that without being familiar with the details of the plant.bob sez he wants to be the new plant manager.

alex sez bob’s superintendents aren’t much good to replace him as production manager, but they both think stacey is good so they offer the job to her and she takes it.

stacey has been examining the queues in front of assembly and in front of bottlenecks - buffers. an item that is scheduled to be worked on for a bottleneck should always be there: if it’s not that’s a hole in the buffer. if there s hole they go to the foreman of the work centre and ask him to work on the hole next.

but they were always visiting the same 6 or 7 work stations - capacity constraint resources (ccrs). the holes have been getting worse ralph sez the ccrs have enough capacity and he’s right on average but if sales increase they’ll have chaos. they need to improve the ccrs too.

stacey wants to be production manager so she work out how to deal with improving ccrs.

ralph noticed that formerly most of his data wasn’t useful for answering the questions alex asked. he wants to develop a system to help reduce the time and effort needed to engineer a sale, to help stacey manage the buffers and to help lou calculate local performance.

34

alex talks to julie about his work and asks her opinion.

julie helps alex realise that jonah cutting him off led to lou coming with the idea of a new measurement system, likewise with the ideas invented by bob, stacey and ralph.

she sez he should ask his team about what techniques are needed for management while he is still at the plant.

alex asks his staff what he should do when he gets to his new job at division.

lou sez he should talk to his new staff for fact finding.

stacey sez he should ask them about their problems and where they stand wrt clients.

bob sez he should find out about local politics, tour production facilities, visit some clients and large suppliers.

alex draws a bunch of shapes on a board to represent the facts he would gather and it looks like a big mess.

alex asks if a committee was trying to work out what was going on how would they do it?

they say the committee would start fact finding.

then they’d try to arrange the facts, and the staff come up with various ways the shapes alex drew could be sorted, but they think it would be a waste of time.

alex sez this idea of arranging the facts does harm in addition to wasting time. companies switch back and forth between arranging things according to product lines or according to functional capabilities. they also switch back and forth between centralisation to reduce duplicated effort and decentralisation to encourage entrepreneurship.

alex sez that the idea of collecting data then arranging it is futile.

35

next morning the staff meet again to try to solve alex’s problem.

ralph sez he had an idea. chemists had discovered a lot of elements like oxygen and hydrogen that other chemicals were made of. mendeleev came up with an idea for arranging the elements. he didn’t use colour, or whether they are gases or whatever.

alex suggests that mendeleev needed a numerical measure, so he could arrange them without being sidetracked by his subjective preferences.

mendeleev used atomic weight. he then noticed that each seventh element had similar chemical behaviour so he arranged them in a table with seven columns. he then spotted holes in the table and used them to predict the properties of undiscovered elements. other scientists were skeptical but mendeleev was right and the relevant elements were discovered, although they found out there was an 8th column: the noble gases.

alex sez they need to look for an intrinsic order in the facts they gather. so they need a technique for doing that and that technique would be useful for management.

later alex tells julie about this and she explains that she went to the library and read some plato.

she heard about the socratic method from alex, who heard about it from jonah so she decided to read about socrates by reading plato’s socratic dialogues.

alex sez that sounds boring. she sez they’re interesting like stories and she’s still reading the first one: protagoras.

36

alex has a meeting at work.

stacey and bob are late cuz they’re dealing with problematic orders.

to manage the division alex has to clarify how he can apply jonah’s ideas in different situations, e.g. - to different plants, to sales etc.

stacey sez they should ask what their goal is as managers. lou sez the goal is to get the division to make more money now and in the future.

stacey sez more isn’t a specific number so that got can’t be achieved.

lou sez she’s right, and they need to move the company toward that goal. to do that they should start a process of ongoing improvement.

nobody knows what this means or how to do it, although hilton smyth uses the phrase “ongoing improvement” in his memos.

bob sez he’s seen a lot of ongoing improvement projects in plants and they’re always a pain in the neck.

bob sez they improved the plant by actually doing stuff rather than talking.

stacey sez the difference between these projects and what they did in the plant is mow they interpreted improvement.

lou sez the relevant measurements are throughput, inventory and operating expense.

most improvement projects focus on operating expenses. lou sez they’ve concentrated more on throughput and inventory, which changed by tens of percent, not operating expense, which changed by 2 percent.

alex wants to know how they’re going to get the division to concentrate on throughput.

bob sez that a process of ongoing improvement has to consist of steps they could do: they need to figure out what process they followed.

ralph sez the process and throughput ideas are connected. a complex organisation is like a chain composed of links. looking at cost is like looking at the weight of the links. but looking at throughput is like considering the strength of links cuz you have use all the links to get throughput. the strength of the chain is determined by its weakest link, which is like the bottleneck.

the process is

step 1 identify bottlenecks

step 2 decide how to exploit the bottlenecks

step 3 subordinate everything else to exploiting the bottlenecks

step 4 elevate the bottlenecks - make them more productive

step 5 look for more bottlenecks if a bottleneck has been broken

stacey sez they went through that process more than once and the nature of the bottlenecks changed.

they started out with production bottlenecks like the ncx-10 and then the market became the bottleneck. she suggests changing the word bottleneck to constraint since calling the market a bottleneck doesn’t make sense.

later alex asks julie how he’s going to continue the momentum.

alex can’t push for more orders cuz that will lead to more bottlenecks and get them back to expediting. but their results don’t justify more staff or machines.

julie sez they’ll have to wait until the plant generates enough money to justify investment.

1 Like

37

alex is in another meeting at the plant.

ralph sez there’s a problem. step 3 sez to subordinate everything to exploring constraints. so when a constraint changes they should change how they operate non-constraints.

the first time a constraint changed was when green tags parts were being released late to the bottlenecks cuz the constraint had moved to how they released work to the plant. this affected the non-constraints.

but they kept using red and green tags to prioritise, but now whatever they release is of the same importance.

stacey has been demanding that stations work on the red tags first and restricting incoming orders. at the capacity constraint resources (ccrs) they work on red tags first and this delays the arrival of green parts. they only notice this when there are holes at the buffer. so they should get rid of the tags.

alex asks is stacey is sure that they can take on more orders.

stacey sez yes cuz this explains why there are holes at the assembly buffer but not the bottleneck buffer. the fact that there are lots of holes means they will have a capacity problem some day, but not now.

did they make a change in on-constraints when the plant’s ability to make money was caused by lack of sufficient orders?

no they even kept releasing material according to the oven and ncx-10 and they were working to 100% of their capacity although they were no longer constraints.

stacey sez that when she realised the load at the bottlenecks was falling she issued orders to put products on the shelves cuz she thought an hour lost at the bottleneck was an hour lost for the entire plant. so they have about six weeks worth of finished goods.

lou sez those can be drawn down but it shouldn’t be done too rapidly cuz it will affect the bottom line.

stacey adds to step 5 but saying you shouldn’t let inertia cause a constraint.

lou sez that they have a new problem. all the sacred cows they had to slaughter came from cost accounting: local efficiencies, optimum batch sizes etc. but stacey had a similar problem with bottlenecks. so changing the constraint means you have to change how you’re working.

alex asks how much load there is on the bottlenecks due to shelf filling and she sez 20%. alex sez this means they can bing more stuff to market so he needs to go see johnny jons with lou and ralph.

38

alex goes to headquarters with lou and ralph to ask johnny jons for more orders.

they’d need over $10 million in additional sales to fill up all their capacity. alex thinks jons can’t have that much business in reserve and they’ll need to come up with new ideas about how to get sales.

jons meets them with dick pashky. jons sez he can’t get them $10 million in new business.

alex sez they can deliver anything in two weeks and their quality is the best in the market.

jons sez sales take time. he has to build up credibility with potential clients.

alex sez he has 20% spare capacity so he needs 20% more sales. he asks jons to consider orders he’d usually decline cuz the quality requirements are too high or the delivery time is too short.

jons sez he accepts any order, but the pressure is too high. lou sez that must mean clients want lower prices.

jons sez he is asked to provide stuff with no marin or below cost.

alex sez he’ll accept orders at 10% below cost. jons sez that makes no sense since they’ll make no money and other clients will start to demand the same prices.

dick sez djangler isn’t connected to their regular customers and they can claim they gave him a volume discount. jons sez djangler wants the $992 model 12 for $701.

alex asks about the materials cost for the model 12 and lou sez it’s $344.07. alex sez he’ll accept the order cuz they’ll get $701 and only pay $344.07 and air freight is less than $30 and he asks to see the details of the deal.

alex points out that cameras from japan are cheaper in new york, but the japanese wouldn’t do a deal where they lose money.

they start working on the deal. alex sez he wants to get into europe so he wants to lock in djangler for one year.

alex some way to compete with european companies other than price. he asks what the shipment time is in europe and jons sez it’s 8-12 weeks. alex sez he’ll deliver any reasonable quantity in three weeks since he already has it in stock.

alex, lou and ralph are happy about the deal, but they think johnny is too attached to cost accounting since he was going to throw away a profitable deal. and division will be even worse, so alex needs to figure out what management techniques to master.

alex thinks that jonah is a physicist and somehow figured out business stuff, so maybe he should read a popular science book.

jons gets the deal and wants to know if he can offer the same 3 week delivery to domestic customers. alex agrees, but he wants to ensure he won’t have to go back to firefighting, so he goes to the library to read science books.

alex reads for some time then asks julie to join him for tea and asks her about why she likes greek philosophy. she sez it’s hard to explain why it’s good and he should read it.

she asks him about science books. he sez physicists don’t start by collecting data. they start with some phenomenon and guess about its cause, then they work out other consequences of that cause and try to verify them and these verifications make it more obvious that the guess about the cause is correct. and they find out that lots of apparently different things have the same underlying cause.

julie sez the socratic dialogues work the same way but they’re about human behaviour.

alex sez business is about materials and human behaviour. so jonah must use these techniques.

julie sez that jonah’s ideas must be about thinking methods not just techniques.

alex sez jonah asked him to figure out what techniques he should learn not how to do them so he’s skipping a step: he should figure out how he operates and how he should operate.

39

alex receives a call from bill peach to congratulate him on his profits. hilton smyth’s indicators are doing well but his plant is losing money.

bill peach sez ethan frost has tried to explain why the indicators don’t produce good results but he doesn’t understand and wants alex to explain it to him at headquarters, and he should come out to find out more about his new job.

stacey sez she needs a meeting because she and bob have been playing expeditors and despite overtime orders will be late.

alex asks if the bottlenecks are overloaded and ralph sez no. stacey points out that the bottlenecks are spending some time idle and then getting overloaded by a wave of work.

they’re doing lots of overtime, but that risks making the plant chaotic and delaying more orders.

ralph sez they should examine a bottleneck but bob thinks they have many travelling bottlenecks. alex sez they should go ahead anyway.

ralph asks what happens if they have a problem at a bottleneck. throughput is lost.

if they have a problem before a bottleneck, then the stream of tasks to the bottleneck is interrupted.

bob sez they always make sure the bottleneck has some inventory.

ralph asks how they know how much inventory they should have in front of a bottleneck.

ralph sez they previously cut the amount and bab suggests increasing it again.

ralph continues to say that if they have enough inventory in front of the bottleneck to last the time required to fix problems earlier, then the bottleneck will keep working.

stacey sez that can’t be true cuz the problems haven’t changed but the bottlenecks are running dry. so they have wandering bottlenecks.

alex sez they should continue to pursue ralph’s line of thought. once the problem is corrected they have to build up stocks in front of the bottleneck again in addition to providing the bottleneck’s standard consumption.

bob sez the fact that they have bottlenecks is necessary cuz if upstream resources don’t have spare capacity they won’t be able to use the bottlenecks to capacity.

ralph asks how much spare capacity they need and stacey sez it’s a tradeoff.

bob sez new orders didn’t create new bottlenecks but did reduce the spare capacity at non-bottlenecks so they need more inventory in front of the bottlenecks.

bob asks ralph to identify the orders that are on short delivery and for them release material one week in advance, for others release it two weeks in advance.

bob asks stacey to get the non-bottlenecks to work through the weekend and not to promise less than four weeks for new orders.

alex is unhappy about what bob had to do. sales will drop their campaign to promise delivery in two weeks, so future throughput will be down.

lou points out that they’ve consumed their whole overtime budget for the quarter.

alex points out that means operating expense is up, throughput is down and inventory is up so everything is going in the wrong direction. they’re also reacting to events instead of planning. alex thinks he must have made a mistake.

1 Like

40

alex has been talking to lou about what’s going on in the division.

bob’s plant has saved the division, it’s the only bright spot. they didn’t realise that until they separated the results for that plant from those from the rest of the division.

alex thinks they made a mistake by just gathering lots of data about the division.

lou sez the information misses lots of details. they don’t have a late receivables report, the information for it is scatted around reports.

lou thinks reducing open receivables by 4 days will help.

alex asks how doing that would improve throughput, inventory and operating expense.

lou sez they’ll be slightly improved and it will give them 4 days of cash. alex doesn’t think this will help.

alex asks how long it will take lou to change the way they evaluate inventory.

lou thinks getting the information will take a few days. explaining it and getting managers to act on it will take weeks.

alex asks how the current evaluation of inventory impacts the finished stocks held by the division.

lou sez sales are down and pressure to show profits is up, so they have built up 70 days of finished products to generate fictitious profits.

alex asks how it changes throughput. lou sez he doesn’t know.

alex asks what reason plants gave for not introducing new models. lou sez they don’t want to obsolete their current inventory cuz it would reduce their bottom line.

alex sez this reduces their market share, and so reduces throughput.

lou thinks they can reduce inventory and improve receivables.

alex asks about plant indicators, judging sales opportunities according to product cost, margins, and anything they can sell above variable cost. and transfer prices between divisions are also a problem. each of these are more than 4 days of receivables.

the next day alex and lou are going through the five steps.

lou asks if they accept the first step is identifying the system constraints.

alex wants to reconsider this idea from first principles.

one principle is that an organisation exists for a reason, another is that it consists of more than one person.

so the organisation requires synchronised efforts of more than one person. so the operation of the organisation depends on chains of links between people, or a grid of chains.

alex sez that the more complex the organisation the more interdepencies there are between different chains and so the fewer independent chains there are.

lou asks what they should do about measurements and sez they are the biggest constraint.

alex disagrees. bigger problems: outdated products, the idea that engineering projects never finish on time, marketing plans that can’t solve the problem, and the idea of blaming other people for problems.

alex thinks all this is a result of some core problem and it must be possible to work out what the problem is.

lou sez that at the divisional level these problems are a result of policies, measurements and procedures, not a bottleneck. alex disagrees: they dealt with bottlenecks in the plant by changing policies, measurements and procedures.

lou asks if there’s no difference why don’t they know what the divisional constraint is.

alex sez the division has excess capacities they’re wasting, not physical constraints and there are markets they could sell to.

alex decides they need a technique to identify constraints, or the core problem. lou sez finding the core problem won’t be enough.

step 2 is exploit the system’s constraint. step 3 is subordinating everything to step 2 but there isn’t a physical constraint so this is unnecessary.

step 4 elevate the constraint involves changing the policy.

alex and lou realise that they need a way to come up with better policies and to test them for whether they solve the problem without causing others.

they need such ideas if they’re going to plan instead of react.

lou and alex realise they need a third kind of idea: how to cause a change of policy.

alex understands now why jonah refused to help him. he has to learn the thinking processes required for solving his own problems.

It would have been good to mention that 40 is the last chapter. Got any concluding thoughts or next steps?

There’s also a long interview with Goldratt at the end of the book, at least in some editions.

I think I understand TOC better now and see more of its relevance to CF. I might do a post on that topic next.

If you decided not to, would you say why?

I haven’t decided not to write the post. I have been busy for the past few weeks. I’ve made some notes today and I’m planning to write the post tomorrow.

Can you see how your reply relates to Curiosity – Question-Ignoring Discussion Pattern ?

Your question is about whether I would say why I decided not discuss TOC and CF to if I decided not to. The question I answered was whether I had decided not to, not whether I would tell you if I decided not to. My response conforms to the pattern:

Me: What do you think about X?

Them: [silence]

Me: Why didn’t you discuss X?

Them: [Starts saying their opinion about X.]

That’s not what I meant. What I meant is that your reply about proceeding with the post is one level of abstraction below what I was talking about.

I was trying to talk at the level of what’s going on. Proceeding with the project (at this specific timing, which seems probably not coincidental but you left that ambiguous) seems possibly designed to avoid that discussion.

So did you want to discuss why I hadn’t written the post and I wrote the post instead of discussing that issue?

I was arranging to move and then actually moving for about the past month. I replied to a few other posts because I could do it easily and quickly. To write TOC and CF relevance material I wanted to reread my notes and work out what to say and I didn’t do that cuz I was busy arranging the move.

You sound maybe defensive. I wasn’t trying to request you do the post now/soon or to make you defensive. Your response seemed like you may have treated my question as a request or reminder to post Goldratt related stuff. I replied to it because it had ambiguity from my pov.