Throughput Accounting 2 · Lesson 3 of 6

The three questions

Every decision comes down to three questions: what happens to throughput, operating expense and investment? Learn TOC’s decision process for capital spending, outsourcing and product viability, and practise it on five common decisions.

18 min · Not completed yet

Replace product cost with three questions

Product cost tried to answer “is this decision good?” by looking at one product. Throughput accounting looks at the whole company and asks three questions:

  1. What is the impact of the decision on the company’s total throughput? (ΔT)
  2. What is the impact on total operating expense? (ΔOE)
  3. What is the impact on total investment? (ΔI)

Then: Δ net profit = ΔT − ΔOE, and the new ROI = (T + ΔT − OE − ΔOE) ÷ (I + ΔI).

That’s all. The skill lies in answering the first question honestly, and for that you need to understand the relationship between the constraint and the company’s products.

Only what changes, only at company level

Look at a decision the way the company’s bank account sees it. Identify the revenues and costs that actually change because of this decision, and evaluate the change at the level of the company, not of a product or a department.

  • Most of a company’s costs are operating expense. They don’t change with a decision unless the decision adds or removes people, shifts, machines or contracts.
  • Truly variable cost changes with every unit, so it always moves with volume.
  • Throughput changes only if price, truly variable cost or volume sold changes.

The decision methodology

Capital spending. Estimate ΔT, ΔOE and ΔI, then ΔROI = (ΔT − ΔOE) ÷ ΔI, or its inverse, the payback period. The crucial rule for ΔT: you can only claim that sales will rise if the resource you are investing in is currently stopping more sales, in other words, if it is the constraint. A faster machine anywhere else gives ΔT = 0, however impressive its efficiency gain.

Outsourcing. Again ΔT, ΔOE, ΔI. Buying in raises truly variable cost on that part, so throughput on it falls. But if the work being outsourced uses the constraint, ΔT also includes the throughput of the extra output the constraint can now produce. That second term usually decides the case.

Product viability, pricing and mix. If there is a constraint, rank products by throughput per constraint unit (T/CU). If there is no constraint in operations, look at throughput per unit. A product with low T/CU isn’t necessarily bad; it may be perfect for times when the constraint has spare time, or for a segment with different prices.

Pause and think: a ₹50 lakh machine would cut cycle time by 30% at a work centre that runs at 70% of its capacity. The vendor shows a 2-year payback from “labour savings”. What is ΔT?

  • Zero: the work centre isn’t the constraint, so no more can be sold
  • 30% of the plant’s throughput
  • The labour savings
  • Impossible to say

And the “labour savings” are real only if people actually leave the payroll, which would show as ΔOE. If nobody leaves, ΔOE is zero too, or positive once you add the machine’s maintenance and power.

Try it

Notice how different decisions that look similar on a cost report come out:

  • The non-constraint machine: ΔT = 0, ΔI and ΔOE up. Every measure gets worse.
  • A second shift at the constraint: a big ΔOE, but a bigger ΔT, because the market is waiting.
  • Cutting stock: ΔI falls a lot, ROI jumps, and cash is released.
  • Cutting people with spare capacity: the only decision that looks good purely on ΔOE, and the riskiest, because that spare capacity is what protects the constraint.

The questions behind every local decision

Go back to the questions from Part 1 that managers answer with product cost. Each has a three-question answer:

Question Throughput accounting asks
Which products should we focus on? Which give the most throughput per constraint unit?
What price should we charge? What does this price do to total throughput, in this segment, without hurting others?
Should we outsource this? ΔT on the part, plus ΔT from the freed constraint, minus ΔOE saved
Is this improvement worth it? Does it let the constraint produce more, or cut OE or I?
Should we buy this equipment? Is this resource the constraint? If not, ΔT is zero.

Check your understanding

Answer all five. Get four right to complete the lesson.

1. TOC’s three questions ask about the impact of a decision on…

  • Product cost, margin and price
  • Efficiency, utilisation and output
  • Throughput, operating expense and investment
  • Sales, marketing and production

At the level of the whole company.

2. You can claim that an investment will raise sales only if…

  • The resource is currently the constraint
  • It reduces product cost
  • The vendor guarantees it
  • It runs faster than the old one

Otherwise ΔT is zero.

3. A decision gives ΔT = ₹40 lakh, ΔOE = ₹10 lakh, ΔI = ₹60 lakh. The payback is…

  • 1.5 years
  • 6 years
  • 4 years
  • 2 years

₹60 lakh ÷ (₹40 lakh − ₹10 lakh) = 2 years.

4. When outsourcing work that uses the constraint, ΔT includes…

  • Only the extra cost of buying the part
  • The change on the part plus the throughput of extra output from the freed constraint
  • The supplier’s profit
  • Nothing

That second term usually decides the case.

5. With no constraint inside operations, product viability is judged by…

  • Product cost
  • Throughput per constraint minute
  • Throughput per unit
  • Total minutes per unit

The market is the constraint; any unit priced above truly variable cost adds throughput.