Industry
Part of Why growth quietly drains cash in most product models
Product models metrics built on contribution per unit
How product metrics built on contribution per unit work: the discount volume multiplier, what a price rise is worth, and realized price against list.
A product business can watch revenue and unit sales every day and still not know whether it is making money on the thing it sells most. The measures that answer that are all built on one number, and it is not revenue.
Every figure below is invented to make the arithmetic legible. Nothing here is a benchmark.
What to take away
- Contribution per unit, meaning price minus the cost that only exists because you made and sold that unit, is the number every other product measure sits on.
- A discount costs far more volume than it looks like it should. The multiplier is the margin.
- A price rise with no volume loss is close to pure profit, which is why it is worth more than almost any cost program.
Contribution per unit
Contribution per unit, worked
- 100list price per unit
- 60unit cost
- 40contribution per unit
- 40,000gross profit at 1,000 units
Two arguments always arrive at this point. The first is which costs belong in the 60. The test is whether the cost exists because this unit exists: materials and the shipment do, the office does not.
The second is what to do with costs that are neither, such as a warehouse that scales in steps. Those belong in a second layer, not in the unit cost. Mixing them makes the discount arithmetic below meaningless. The full version of that argument is in unit economics.
What a discount actually costs
Hold gross profit constant and ask how much extra volume a discount requires. With a margin of m and a discount of d, the answer is d divided by (m minus d):
Volume needed to hold gross profit
5% off
- 30% margin
- +20%
- 40% margin
- +14%
- 50% margin
- +11%
- 60% margin
- +9%
- 70% margin
- +8%
10% off
- 30% margin
- +50%
- 40% margin
- +33%
- 50% margin
- +25%
- 60% margin
- +20%
- 70% margin
- +17%
20% off
- 30% margin
- +200%
- 40% margin
- +100%
- 50% margin
- +67%
- 60% margin
- +50%
- 70% margin
- +40%
In absolute terms on the example above: at 10% off, the price is 90, contribution is 30, and you need 1,333 units instead of 1,000. At 20% off, contribution is 20 and you need 2,000 units, double the volume for the same gross profit.
Print that table. It takes a minute to compute for your own margins and it is the single most useful thing to have in a room where discounts get agreed. Where the discounting is structural rather than occasional, the problem is the price structure itself, which is the subject of pricing architecture.
The same arithmetic in reverse
A price rise with no volume response is worth more than it feels like. On a book of 1,000,000, adding 2% is 20,000 and all of it is gross profit. Adding 5% is 50,000.
To make the same 50,000 from volume at a 40% gross margin, you need 125,000 in additional revenue. That is two and a half times the selling for the same result.
Price is not a lever like the others, and treating it as one is the most expensive habit in a product business. Formal price theory is built on this asymmetry, and MIT's pricing course is organized around it.
Realized price against list price
List price is a claim. Realized price is total collected divided by units shipped, and the gap between them is made of discounts, rebates, promotional credits, returns, and freight you agreed to absorb.
Compute it monthly, per product, and per channel. Two things usually turn up the first time. One product's realized price is far below list because a single large account has an arrangement nobody remembers agreeing. And one channel's realized price is lower than management believed, because the deductions are booked somewhere other than sales.
Where the same product goes out at different net prices to competing business buyers, the arrangement has legal edges as well as commercial ones, and the price discrimination provision is where those are written down.
Measures worth keeping beside those
- Contribution by product, sorted.Almost always fewer products carry the margin than anyone expects.
- Return rate and the cost of a return.A return is not a reversed sale. It is a reversed sale plus freight, inspection, and often a write-down.
- Inventory turns.Slow inventory is cash you already spent sitting in a room, and it does not appear as a loss until you mark it down.
- Share of units sold at full price.If it is small, your list price is a fiction and your structure needs redesigning rather than defending.
What each of these implies for the wider shape of the business is set out in product models, and how the money is triggered in the first place belongs in revenue models.
Common questions
Should freight be in the unit cost?
If it moves with the unit, yes. Firms that leave it out get a contribution figure that looks fine and a bank balance that does not. The test is not accounting convention, it is whether the cost disappears when the unit does.
Our costs move monthly. Which cost do we use?
Whichever one you can defend, stated with a date. A contribution figure computed on a cost from eighteen months ago is worse than no figure, because it is believed.
Is gross margin percentage or contribution per unit the better measure?
Per unit for a pricing decision on that product, percentage for comparing across products. Mixing them is how a firm ends up promoting the high-percentage product that contributes very little cash.
How often should the discount table be recomputed?
Whenever your gross margin moves by more than a couple of points. It is a two-minute calculation, and a table built on last year's margin gives permission for discounts that no longer hold.
