You get out of a price war by moving first, moving on a narrow front, and moving visibly. Raise a small, deliberately chosen set of lines, hold the new level long enough for competitors to observe it, and explain the change to customers in terms that have nothing to do with any competitor. The customers you are most afraid of losing are rarely the ones buying the line you discounted.
The uncomfortable part is that the exit is a unilateral act. You cannot coordinate it and you must not try. What you can do is make your own behaviour easy to read, so that a competitor who would also like to stop has something to follow.
Why price wars carry on after everyone wants them to stop
A price cut is visible within a day. The volume response takes weeks. The margin damage appears a quarter later in a report almost nobody reads on time. So the decision loop runs on the fastest available signal, which is the competitor's price, and the slow signals never get a vote.
Two other mechanisms keep it running. Attribution: your rival cut on Tuesday, you assumed it was aimed at you, and it was an overstock clearance in one warehouse. Ratchets: discounts get built into customer expectations, sales targets and channel rebates, so undoing one cut costs more internally than making the next one.
Is it a price war, or one competitor clearing stock?
Spend a week establishing which of the two you are in. The correct response differs completely and the diagnosis is cheap.
- Breadth: how many lines moved. A war is broad and repeated. A clearance is narrow and ends when the stock ends.
- Pattern over time: three cuts in four weeks on the same lines is a war. One cut that then sits flat is a new price level.
- Stock signals: check availability, delivery times and pack sizes on the cut lines. Items that go out of stock at the low price were being cleared.
- Who moved: one regional seller, a marketplace reseller, or the two competitors who actually set price perception in your category.
- List price or promotional price: a promotion with an end date is a campaign, a changed list price is a position.
If it is a clearance, the right response is usually nothing, plus a note in the calendar to look again in three weeks.
The narrow front: which lines to raise first
Do not raise everything, and do not start on the lines the war is being fought over. Start where an increase is least likely to be read as weakness and least likely to cost volume.
- Lines with low price visibility, where pack size, specification or service content differs across sellers and a buyer cannot compare in under a minute.
- Lines where you hold a real availability advantage this month. Availability beats a two per cent gap for most business buyers.
- Lines already losing money at the current level. The worst case is that you lose volume you were paying to have.
- Lines bought alongside several other items. The basket protects you.
- Never the three or four lines your customers quote back to you from memory. Those hold until the rest has survived a month.
Move by an amount large enough to be seen and small enough to survive. On a category sitting at index 96 against the main competitor, going to 99 or 100 is a signal. Going to 96.5 is noise.
The arithmetic to run before any defensive cut
Take a line at 30 per cent gross margin selling 100 units a week. Weekly gross profit is 100 times 30, so 3,000. Cut the price by 5 per cent and unit margin falls from 30 to 25. To hold 3,000 you need 3,000 divided by 25, which is 120 units: 20 per cent more volume from a 5 per cent cut.
Repeat it at your real margins. At 20 per cent, a 5 per cent cut takes unit margin from 20 to 15, and 20 divided by 15 is 1.33, so you need a third more volume. Most defensive cuts get approved without anybody stating the volume forecast they are implicitly signing up to.
What you tell the customer
The message that works is specific, forward looking and silent about competitors. Buyers accept an increase attached to something they can picture: an input cost, a stock commitment, a delivery window, a longer warranty, a fixed price period.
Two things to avoid. Do not apologise, because that invites negotiation. Do not promise this is the last increase, because you do not know that and it will be quoted back to you.
Give large accounts notice before the change appears on the site or in the price list. Notice costs nothing and removes the usual reason a buyer escalates, which is being surprised in front of their own manager.
Signalling in public, legally
Everything here is unilateral and public. You publish your own prices, you explain your own reasoning to your own customers, and you let competitors observe the result. What you never do is contact a competitor about prices, discuss future pricing at a trade association, or pass intentions through a shared supplier or distributor. In the EU and the UK that is a serious infringement and no margin recovery is worth it.
Signalling in public means holding the new price for a defined and visible period. A price that slides back within ten days teaches the market that your increases are negotiable.
If nobody follows
Decide the fallback before you start. Set a window, four to six weeks for most weekly buying cycles, and decide in advance what you will measure: unit volume on the raised lines, gross profit in currency rather than margin percentage, and whether any competitor moved at all.
If volume holds within a few per cent, keep the level and widen the front. If volume falls hard on lines you thought were hard to compare, you have learned they are more comparable than you assumed, which is worth the cost. If a competitor undercuts the raised lines immediately and specifically, you are facing a deliberate strategy, and the honest answer is to hold and compete on availability, terms and service.
The failure mode to avoid is the partial retreat: raising prices, losing nerve in week two, and quietly reverting half the lines. That buys the volume loss and none of the margin.