Clarifying the Concept of Throughput

Throughput (T) is a central concept in TOC and it should be a central concept for all management. I see it as expressing the added-value created by the organization. I focus on the value of using T, I and OE for making superior decisions.

The formal definition of T for profit organization is: Revenues minus the truly-variable-costs (TVC).

Don't forget the cost of the water when predicting the T
Don’t forget the cost of the water when calculating the T

We use T as a periodical performance measurement and use it also for judging the economic impact of a single decision – the delta-T generated by the decision minus the delta-OE (operating expenses) caused by the decision at hand.

The revenue part of the definition of T is relatively straight-forward. The definition is still open whether one should consider the timing of receiving the revenues. Is T of $1,000 to be received next year the same as T of $1,000 today? This is quite a topic to deal with at another time.

An even more relevant issue is the impact of uncertainty on the revenues. As decisions always aim at the future, the prediction that we’d generate revenues is always based on forecasting.

On top of how to predict the revenues there is another issue: what should be included in the TVC? Eli Goldratt added the term “truly” to the “variable cost” to warn us to be careful about what we regard as TVC.

I like to highlight two important problematic areas with TVC.

One: are the costs of raw-materials always TVC?

Two: high level decisions might include costs that do not vary with smaller decisions. Should we treat those as TVC?

Why do we claim that raw materials are truly-variable-costs?

Purchasing raw materials is usually done long before firm orders for the end-product are received. It means that the cost has been spent no matter whether we have sold or not. Thus, the cost is not directly triggered by the sale! So, why do we usually consider the material cost as part of the T and not part of the operating expenses (OE)?

When raw materials are regularly replenished we can safely assume that any sale of end-products initiates purchasing of materials.

In such a case it is right to treat the cost of raw materials as TVC as we are used to do.

This observation means that when a decision to stop further production of specific items has been made then the subsequent sales of that product should not treat the cost of materials as TVC. It is a mistake to refrain from selling items, whose production is stopped, because the market price is lower than the cost of materials. The only two relevant questions are whether we have enough available capacity, which cannot be used for existing demand, and whether selling old products for reduced price would cause loss of sales of full-price items or reduced the image of the brand in the market.

Suppose you hear that the cost of materials went up by 20%. Fortunately you still have enough inventory of the materials for the next two months. When would you update the TVC, causing significant reduction of your T per product-unit?

My view is that the T should be updated immediately!  Remember the assumption: any sale triggers replenishment of the materials. This is the basic assumption behind including the material cost in the TVC.

The other case to consider is the variable costs of a high level decision. Example: The TVC of adding a passenger on a flight is pretty low, mainly the impact of the additional weight on the fuel consumption. However, when a decision to cancel, or add, a flight is considered we are exposed to very different level of variable costs. The fuel consumption of the whole trip is definitely a significant variable cost item. The cost of the crew include variable elements that depend on actual hours of flight and the stay in a far away city. I claim that even the maintenance costs of a flight have to be considered TVC, because those costs are mandatory for X hours of flights, thus any cancelation of flight reduces the relative future cost of maintenance.

However, should those “higher level TVC” be part of the T?

I suggest calling TVC only the truly-variable-costs of a SINGLE SALE! For such a sale we have a clear definition of throughput and any high level decision, like adding or cancelling a flight, would anyway involve certain delta-OE, which considers all the expenses caused by the decision. Thus, the decision criteria of delta-T – delta(OE) > 0, is still intact.

Readers who happen to live in Europe and look to understand the potential of throughput accounting for their business, and to be exposed to wide expansion of the power of T, I and OE, should be aware of my session in Paris in October. Have a look at:


Published by

Eli Schragenheim

My love for challenges makes my life interesting. I'm concerned when I see organizations ignore uncertainty and I cannot understand people blindly following their leader.

16 thoughts on “Clarifying the Concept of Throughput”

  1. Eli, I find it helpful in decision making to view TVC to be included in the Throughput calculation as those expenses that will change without a doubt as a result of the decision. The expenses that are always necessary to have the sale, mandatory for any change in Sales Revenue. For this reason I like to use the term True Variable Expenses (TVE) to provide clarity. Also because the term “costs” have so many other meanings.


    1. Hi Philip, I fully agree that TVE is a better expression than TVC. I use TVC just because it seems more popular. I also dislike the title of Throughput Accounting, because I think it has very little to do with Accounting, but I use it as it is a common name.


  2. There are such raw materials that I would like to finish (get rid of). I suggest not to include them in TVC since I have no intention to replenish them.


    1. If you can tie the purchase of RM to a specific product and it is used immediately then I would consider the cost TVC. However this is the exception, normally RM is purchased in bulk and should not be captured as TVC. In addition there are many product combination of RM that is used immediately and RM that is not. This creates an accounting nightmare if all of the RM is carried as TVC or even as OE.

      However if the RM purchases are carried as I (Investment or Inventory) Goldratt used I to describe both. Then it seems to me that the RM (I) becomes (TVC) when used by a product. As an example: A shipyard purchases 25 tons of steel for 5 ships they are building. Now you may be able to purchase the steel in smaller batches, however the steel is transported by train and there is a long lead time. So this purchase is captured as I and then converts to TVC as it is being used. Now this is a large product, however any solution has to be scalable for large and small products.

      In the airline industry fuel is a large cost driver and and an airline will commit to purchasing in bulk to guarantee availability and to lock in a favorable price. Again my suggestion is to capture as I and convert to OE (at the Business Unit Level) not TVC at the product level. There are several reasons for this. You do not know how much fuel an aircraft will actually use on any given flight since the fuel consumption has multiple influencing variables such as weather, aircraft traffic, winds, amount of reserve fuel required. In addition the fuel in aircraft may become scrap because of fuel contamination and in some cases due to maintenance requirements. You do not accurately track the delta between the amount of fuel pumped into the aircraft and what was actually used.

      I would suggest capturing all of the aircraft maintenance costs as OE. The costs captured are based on flight hours, cycle time of components, unscheduled and scheduled maintenance (much of the cost is based on calendar requirements not usage). Components are routinely moved from one aircraft to another and in some cases must be removed and replaced with a new component. Therefore the maintenance costs associated with usage and or cycle times moves from one aircraft to another, this is why I believe OE at the BU level is the best pace to capture the costs

      I apologize for the length of the comments but this is a very important topic, using delta T for decision making, we must make sure it covers every scenario before we put in the TOC toolbox.


  3. I think we should clearly distinguish between accounting, which has to capture the movement of money and where it is stuck, and decision making.

    When bulk of materials are purchased a good way to capture it is to treat it as ‘I’. Well, actually I don’t care because this is an accounting problem how to register the purchase and it is not my expertise.

    What I do care is to make my decisions right. So, when I produce a product, hoping to sell it for $100, should I consider the cost of materials, say $75? After all I have already paid for those materials, and I have also paid for the purchasing of the machines and the whole building. But, while the cost of the machines would not be changed by producing another product-unit, certainly the cost of the building would not change, using materials from stock would, most of the time, cause another purchasing of those materials. This happens when the RM are in constant use. I don’t care whether they are used by many end-products or only for one end-product, the point is: any use of materials somehow, not necessarily immediately, will be reflected in a future purchase.

    But, if I have RM that would not be used anymore – then I might just find a use for whatever is left and no future purchase will occur. Suppose that the shipyard Danny has mentioned eventually built only four ships and the order for the fifth was cancelled and no more ships of that kind would be ordered in the foreseen future. Suppose also that part of those materials cannot be used on the newer line of ships which might be ordered in the future.
    Suppose now there is an idea of building yachts using the same materials, which cannot be used for the newer line of ships, and trying to sell them cheaply, as the competition is particularly strong. So, the whole potential T from producing such yachts, by a shipyard that has not been at that market segment, is based on getting rid of those materials for part of their formal cost.
    Should we include the historical cost of those materials in our decision? I believe that this is what Utkan had in mind and I agree: The T of the yachts should NOT carry the cost of materials I do not intend to buy any more of. I let the accountants to find the way to write-off the investment in those materials.


  4. I agree, obsolete materials should be written down to 0,- value and therefore would contribute 0,- to the TVC of a decision to utilize them. Same would go for Minimum Order Quantity (MOQ) materials, which is something I have experienced often in projects. Projects will have a budget for and possibly also order intake compensation for any excess materials due to MOQ. What often happens thou is that the value of the excess parts is never written down or expensed to the project, meaning the project over performs and the excess parts stays in inventory at full value, causing bad decisions to be taken later, due to the costs being taken into consideration and finally down the line an auditor points out that the item is obsolete and has to be written down, possibly at a stage where utilizing it for anything is no longer possible.


  5. Hi Eli,
    You thought do not treat RM costs as TVC when production had been stoped puzzled me very much. I tried to catch an idea almost whole night.
    So what conclusion did I came to?
    First of all there is two puporses which seem different at a first sight: accounting and decision making. But I think that multipying tools is not good idea.
    If we consider T-conception separatly, accounting RM costs for items which are not produced causes flawed decision about the benefit of saling at a reduced price. But..
    Let’s remember that if we don’t consider RM costs as TVC it causes distortion in CF evaluation.
    If I’m not mistaken, the formula for SINGLE SALE is CF = Q*(P-TVC)- delta OE-delta I.
    Example 1: if we replenish RM as much as sold: CF=1*(100 – 50) – 0 – (50-50) = 50, equal T
    Example 2: if we don’t treat RM as TVC and don’t replenish and sale at reduced price: CF = 1*(80-0)-0-(-100) = 180 (?) even more than revenue, How could it be?
    Example 3: if we treat RM as TVC and don’t replenish and sale at reduced price: CF= 1*(80 – 100) – 0 – (-100) = 80, cash flow is equal to revenue although T is below zero. DEcision is obvious – to sale. Here we don’t make a lot of changes in accounting and we don’t look on T separatly – we look at a whole system. And we have grouds for correct decision.
    What do you think?
    P.S.: Sorry for possible mistakes: reading and speaking English is easer than wrighting 🙂


    1. I don’t think you can include Delta I in the net profit calculation, but would have to include it in the Return on investment calculation.


  6. Another issue when using throughput accounting, is how it handles decisions in a group environment, where companies owned by the same legal entity is making transactions with each other. If you bought the raw materials from your sister company, how much of those costs are TVC? I guess the correct answer would be only the TVC your sister company incurred to make that sale, correct?

    How about changes in OE when faced with a decision.. do you only take into consideration your own delta OE or how about your sister company’s change in OE if you decide to accept the order? We know the answer of course, but to get that into the decision making process in a group environment is a challenge!


    1. Rasmus, it depends from what perspective you look. If I have to justify my own company, then I’d treat the sister company as a regular supplier. If the global company likes to have superior decisions and adjusts its performance measurement to reflect the holistic considerations then the transfer between the sisters does not matter: you just focus on what you get from outside the company and what TVC you pay to parties outside the company, no matter which sister paid it. Same for delta(OE). For that matter, overtime paid to employees is a payment outside the company (paid to an individual).


  7. Dmitry, I admit to be confused by your examples. In example 2 you write: CF = 1*(80-0)-0-(-100) = 180. I have no idea what is CF (is it the same as T?) and what are the numbers. I remember using an example where the sale price is 100 and the materials 75.

    On the conceptual level Rasmus is right. The decision criteria is delta(T) minus delta(OE). When delta(I) is involved then you need to compare the stream of T to delta(I) to check the ROI. Another way is to transform the delta(I) to a stream of delta(OE).

    Anyway, without understanding what your numbers stand for I cannot really comment much.


    1. I don’t speak about Net Profit. Profit is not the goal. The goal is to make money. I speak about making decision and cash flow.
      About numbers: price is 80, raw material cost 100 delta OE is zero. So, when we sale product unit for 80 we have loss for 20 if we consider RM as TVC and profit 80 if we consider it zero. So, using this RM reduces our Inventory for 100. Money stuck in the system can’t disapear.
      If we don’t consider RM as TVC what do we have to do whith our I: we have to make it zero too? How we’ll balance vaporized money? I’m looking at a whole accounting sub-system and its behavior.
      If we say that our goal is “to make money” (not profit, profit isn’t money at al, I think difference is obvious) when we sale and replenish RM we can’t sale with negative T at all.
      But when we sale and don’t replenish we only return stucked money which we can not return other way. As much as we can and profit doesn’t matter at all.
      If the goal is money all we account in “I” is very important.
      That’s why I think that we always have to consider RM as TVC.


  8. Dmitry, first if your example 3 you have reduced the cost of materials twice. Anyway you should not include the delta(I) for the T.
    Once the money for an investment is spent – the money is gone. When you like to judge your decision to invest you go for ROI. Investments might be fully lost, this is not a rare incident, like having to scrap materials, or even equipment. So, there are accounting ways to report it.


  9. Hi Eli,
    I’m trying my best to understand your doubt about consideration RM costs as TVC in case if we sale goods which we don’t produce else. The issue is very important for me and I’m going to make it as clear as possible. Sometimes it seems to me we discuss it in a style like:
    “- Cat is seating.
    – No! Cat is wight!”

    Two questions which I want to be clarifyed:
    1. What is the purpose not to treat RM costs as TVC if we don’t replenish RM?
    2. What changes have to be done in information and accounting system to support this purpose?

    My first assumption: “making profit” is not the same as “making money”. There are a lot of ways to have very good profit but have very lack or absence of money. So, “making money now and in the future” is not the same as “making profit now and in the future”. Profit is nesessary but not sufficient condition for making money. That’s why I claim that profit is not the goal of organization. If the goal is to make money then main measurement is not Net Profit!!! The measure is Free Cash – money which ouner can use to any purpose he wants with out any harm to organization. I saw many entrepreneurs who didn’t distingiush profit and free cash and it always was great problem to their companies. Profit correlates with free cash in long term conditions especially if we continue to replenish RM without making stock excess. But it’s very low related in cases when we stop replenishment.

    That’s why I think that all three elements – T, OE and I have to be taken in consideration when we are making decisions. When you speak about Net Profit you need to take in consideration only delta T and delta OE. But if we are speaking about MONEY delta T and delta OE is not sufficient. We need to take in discussion delta I. Because the impact on free cash may be expressed in the next way: delta CF (or free cash) = delta T minus delta OE minus delta I. Especially if we are talking about single sale.

    The difference between Net Profit and free cash is not very inportant when delta T > delta OE and T is above zero then it’s only the issue of investment handling – not to put all our profit to the stock. But it becomes very important when we sale products below RM costs for products stoped producing.

    What happens when we have any single sale? Let put aside OE for a moment. We get some revenue, we fix TVC and we reduce our inventory as much as TVC we fixed. If we don’t replenish nothing else happens. That’s why in example 3 I’ve reduced RM costs twice: first as TVC reducing T and second as delta I reducing inventory. The result is free cash equal to revenue and negative Throughput and Net Profit. And I think it’s correct because it means that we lose money invested in past (80-100=-20) but we have real money on hand (80) instead of laying on our stock may be for eternity.

    So, in example 2 I had no ideas where I have to account RM costs if not TVC that’s why the result became unreasonable.

    I think if we base on assumption that the goal is “make money” (not profit) is valid than TOC already have tools for making superrior decisions and there is no need to doubt if RM costs are always TVC because it causes more complexity than benefit.

    Nevetheless, question about point of accounting revenue and T in time still exists.


  10. Dear Mr. Schragenheim,
    we are in the process of implementing TOC in one of our constructions sites.
    On the reporting side we came across the following question:

    “Throughput” is the name given to the money that is generated from sales – which means the value of sales less the TVC – Truly Variable Cost – only the cost of making and selling additional units.

    There is two ways of deducting TVC from Sales.

    You can take sales, let’s say from the month of August 2016
    and deduct the TVC which occurred in August 2016.


    You can take sales, again from August,
    and deduct the variable cost which were used to make the products
    which are sold in August.

    So, the question is: do I use a period based calculation or a
    product cost calculation when defining the TVC.

    For example:
    You could have a company-wide holiday shut-down for the entire
    month of August, how it sometimes happens in Italy or France.
    However, your sales people stay in the office and sell some products
    from stock.
    I guess it only makes sense if you take the product cost which occurred
    when producing the products (a couple of months ago) which were sold.
    Otherwise your entire Sales would be the same as Throughput,
    as there is no cost in production for August due to the shut down.

    If you could answer this question I would very much appreciate.

    Thank you


    1. Very good question Andreas, and you have answered it correctly. Before there is a sale the cost of the materials that go into production is actually an investment (not from pure accounting point-of-view they look on investment only when the return would be in later years): an expense with the hope it’d turn into sales. Once sales are materialized the TVC, done quite some time ago, as you said, becomes part of the throughput. This is the regular view of TOC so far. In the post I claim that the rational of treating the cost of materials as TVC is that the sale would cause that expense in the near future: trigger purchasing the materials in order to replenish what was sold. Looking at it from this angle makes a causality linkage between the decision to sell for a price, and being aware that decision would lead to the expense of buying additional materials.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s