A healthy overall margin is compatible with a third of the range losing money. The aggregate figure cannot tell you which third, and pricing decisions made from it are guesses with a decimal point.
Gross margin is the number most owners can quote from memory. It is also an average, and an average is a single figure standing in for a distribution it tells you nothing about.
A business with a thirty per cent blended margin might have that margin on everything. Far more commonly it has fifty per cent on some lines, fifteen on others, and a handful sold below cost that nobody has identified because they have never been looked at separately.
Why loss-makers survive
- They were priced years ago against a cost that has since risen.
- They carry a discount agreed for one order that became the standing rate.
- They are high-volume, so they look important in revenue terms.
- They belong to a large customer, and nobody wants to raise it.
- They consume disproportionate service, handling or credit that never lands against them.
Nothing in an aggregate margin identifies any of these. The figure can be improving while the mix beneath it deteriorates.
Contribution, not gross margin
The more useful measure is contribution: revenue less the costs that actually vary with that product or customer. That means getting costs to the right place rather than leaving them in a pool.
- Build a proper landed cost
Purchase price plus freight, handling, duties and the cost of holding the item. A margin computed against invoice cost alone flatters everything that is slow-moving or expensive to move.
- Attribute the variable service cost
Delivery, installation, returns handling, technical support. A customer who orders weekly in small quantities costs more to serve than one ordering monthly in bulk, at the same price.
- Charge the cost of credit
A customer on ninety days is being lent money. That has a cost, and it belongs against their contribution.
- Deduct what is given away after the invoice
Schemes, credit notes, rebates and returns, matched to the line that generated them rather than absorbed centrally.
- Then rank
By product, by customer, by channel. The ranking is usually more surprising than any single number in it.
What tends to come out
Two patterns recur. The first is a concentration of profit — a minority of the range or the customer list producing most of the contribution. The second is a tail that is not marginal but genuinely negative, where more volume makes things worse.
Neither is a crisis. Both are decisions: reprice, restructure the terms, change the service model, or accept the loss knowingly because the line does something else for the business. The point is that it becomes a decision rather than something happening by default.
On accepting a loss-maker deliberately
There are good reasons to keep one — it completes a range, it holds a customer relationship that is profitable overall, it absorbs capacity that would otherwise sit idle. Those are legitimate and they are strategy.
What is not legitimate is not knowing. The difference between a subsidised line and an undetected one is whether anyone chose it.
Where to start
Not with a costing system. Start with one month, the top twenty products or customers by revenue, and honest attribution of the costs that vary. That is usually enough to show whether the distribution is tight or wide — and if it is wide, the case for doing it properly makes itself.