Hello TOC

Finance & Measurements, as Taught by Eli Goldratt · Lesson 5 of 8

The sacred cow of product cost

Why transfer prices punish improvement, and why “product cost” is the most cherished and most damaging concept in industry. Then Goldratt's simplest proof: a plant with two products, accurate data and no excuses, where following product cost loses money.

22 min · Not completed yet

Profit centres and transfer prices

Many companies judge their subsystems as profit centres: each department or plant gets its own profit and loss statement. When a profit centre sells to another part of the same company, the price is a transfer price, usually set from cost:

direct labour × overhead factor + raw material + a % margin = transfer price

Pause and think: a profit centre becomes more efficient and its costs fall. What happens to its profitability?

  • It rises
  • It falls: lower cost means a lower transfer price and margin
  • Nothing changes, since transfer prices are set once a year

And if its costs rise, its transfer price rises and so does its profit. Transfer prices based on cost demotivate improvement. Goldratt returned to profit centres in his supply chain session; lesson 8 covers his conclusion.

The most cherished sacred cow

“Product cost” is probably the most cherished concept in industry. The idea is simple: as long as the selling price is higher than the product’s cost, the company makes money by producing and selling it.

Goldratt asked an awkward question. Take almost any company that is losing money. Look at its products. For nearly every one of them, the price is higher than the product cost. If the concept were right, how could that company be losing money?

The usual excuse is that the data wasn’t accurate or up to date. Goldratt’s answer: the problem is not in the data; it is in the concept.

A proof with no excuses

To show it, he used an extremely simple example, with every excuse removed. All the data is fixed and accurate:

  • The plant makes only two products, P and Q.
  • The time each resource needs for each step is known and doesn’t change.
  • The market demand is known: 100 P and 50 Q a week.
  • Selling prices don’t change: P sells for $90, Q for $100.
  • Raw material prices don’t change.
  • Only four resources, A, B, C and D, one of each, each available 5 days a week, 8 hours a day: 2,400 minutes. No overtime, no second shift.
  • Operating expense is known and stable: $6,000 a week.

Before reading on, take a pencil and answer one question: what is the maximum net profit (or minimum loss) this company can make in a week?

Is there enough capacity?

Add up each resource’s minutes for the full demand. A needs 2,000, C and D 1,750 each, but B needs 3,000 minutes and has 2,400. B is the bottleneck: the company can’t make everything, so it must choose which product to make in full and which to make with the time that’s left. (In reality, Goldratt warned, you don’t find the bottleneck by calculation: data inaccuracies lead you astray. Here the data is perfect.)

The product-cost judgment

Product Q has the higher gross margin (price minus raw material: $60 against $45), and it uses less labour (50 minutes against 60). By product cost, Q is clearly the product to push. So: make all 50 Q, then P with the time left. Play it:

Following product cost, 50 Q and 60 P, the company loses $300 a week. Violating it, 100 P and 30 Q, it makes $300 a week.

Sacred cows make excellent steaks.

Product cost is arbitrary. It doesn’t reliably point to the products that increase the company’s profit, and it can promote products that reduce it. So how should we decide which product to promote? That needs a different way of looking at the company, which is the next lesson.

Every judgment, one cause

Look back at the five judgments. The financial statements reward building stock. Equipment justifications count savings that never appear. Make-or-buy outsources the wrong parts. Transfer prices punish improvement. Product cost picks the wrong products. In every case, the distortion comes from the same mechanism: cost allocation.

Check your understanding

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

1. Cost-based transfer prices…

  • Reward improvement
  • Have no effect on behaviour, since they cancel out in the group
  • Demotivate improvement: lower cost means lower profit for the profit centre
  • Are required by law

Higher cost, higher price, higher profit.

2. Goldratt’s awkward question about loss-making companies is that…

  • Nearly all their products are priced above product cost
  • Their data is always wrong
  • They have no products
  • They don’t calculate product cost carefully enough

So the concept, not the data, must be wrong.

3. In the P&Q plant, the bottleneck is…

  • A
  • C
  • D
  • B, needing 3,000 minutes against 2,400

So not everything can be made.

4. Following the product-cost judgment (Q first), the company…

  • Makes $300 a week
  • Loses $300 a week
  • Breaks even
  • Makes $1,887 a week

Making P first makes $300.

5. The common cause of the distortions in all five judgments is…

  • Late data
  • Bad software
  • Cost allocation
  • Inflation

Which raises the question: why do we keep using it?