Loss leader: selling below cost on purpose, and when it works

A loss leader is a product deliberately sold at or below cost in order to attract customers who will then buy something else at a normal margin. Supermarkets are the textbook case — milk and bread near the back, priced to pull people through the door and past everything else. The logic is sound and the execution is where it usually falls apart, because the whole thing depends on a second purchase that nobody has verified will happen.

It is worth separating from ordinary discounting. A sale price is a temporary reduction on something that is normally profitable. A loss leader is a permanent, structural decision to lose money on one line and recover it elsewhere.

When a loss leader works

The strategy has four preconditions. Miss any one and you are simply selling something at a loss.

• The follow-on purchase is close to inevitable — the printer needs ink, the razor needs blades, the basket needs the rest of the shopping.
• The profitable product has enough margin to cover the loss with room left over.
• The customer cannot easily buy the cheap item alone and go elsewhere for the rest.
• You can measure the pairing, so you know whether the second purchase is actually happening.
• The loss-leading item is not so easily resold that buyers stockpile it.

The failure mode has a name in retail — cherry-picking. Customers buy the underpriced item, nothing else, and leave. Online this is not an edge case but the default, because comparing prices across shops costs nothing and buying only the discounted line is trivial. A loss leader that worked in a physical shop frequently does not survive being put on a website.

The versions of it that mostly fail

In services and software, the loss leader usually appears as free or heavily discounted onboarding, a first project at cost, or an introductory rate. The intention is to buy a relationship. The result depends entirely on whether the discounted first engagement resembles the profitable ongoing work, and often it does not: setup is intensive, ongoing work is light, and the customer has been trained to expect a price that was never sustainable.

The other common failure is anchoring. Whatever price a customer first pays becomes their reference point, and the increase to the real price reads as a rise rather than an end of promotion, however clearly it was labelled. If the gap is large, expect to lose a meaningful share of those customers at exactly the moment they were supposed to become profitable.

Deciding whether to try it

1. Work out the true cost of the loss-leading item, including delivery, payment fees and support time.
2. Identify precisely which product recovers the loss, and calculate its margin.
3. Estimate what share of loss-leader buyers must buy that product for the pair to break even.
4. Compare that share against what you actually observe today. If you have no data, run it small first rather than across the range.
5. Check that the item cannot be bought alone and cheaply resold or stockpiled.
6. Set a review date and a stop rule before launching, not after.
7. Track the pairing per customer, not the totals — aggregate revenue hides the fact that two different groups are buying.

Loss leader versus penetration pricing

They are frequently confused. Penetration pricing sets a low price across the whole offer to win market share quickly, with the intention of raising it later. A loss leader keeps normal prices everywhere except one deliberately chosen line. Penetration pricing bets on the market; a loss leader bets on the basket. The risks differ accordingly — penetration pricing risks never getting the price up, a loss leader risks the second purchase never arriving.

There are also places where selling below cost is restricted. Several jurisdictions limit below-cost selling, particularly in food retail and where it looks like predatory pricing by a large player, and some regulate how introductory prices must be presented. Check local rules before building a strategy on it — this is a matter for a lawyer in your market, not for a blog post.

Measuring it honestly

The only measurement that settles the question is per-customer: of the people who bought the loss leader, how many bought the profitable product, how soon, and how much did they spend in total. Revenue and volume both go up when you underprice something, which is precisely why looking at either one will tell you the strategy is working when it is not.

If you invoice rather than sell through a till, this is more visible than it seems. Ettex Invoices records what each customer bought and when, so the pairing can be read off the customer history rather than estimated — did the client who took the discounted first job come back for the full-price work, or did they take the cheap piece and stop.

It is an invoicing and billing tool, not a pricing engine: there is no basket analysis, no elasticity modelling and no automatic margin alert. It gives you the record. Deciding what the record means, and whether to stop, is the part that has to stay yours.

A loss leader is a tactic inside a pricing strategy, not a substitute for one. If the surrounding prices were never set deliberately — competitive pricing on some lines, cost-plus pricing on others, no stated logic anywhere — then the loss leader is simply one more unexplained number, and the margin it is supposed to be buying is not being measured.

A loss leader is a bet that the second purchase happens. Calculate the share of buyers who must make it, measure per customer rather than in total, and set the stop rule before you launch.