The Price War and Why the Second Mover Loses More

Why this matters

A competitor drops their price, you drop yours to hold the work, and six months later both shops are doing the same volume at a worse margin with the customer's expectations permanently reset. That is the normal outcome, not the unlucky one, and the reason is structural rather than a matter of nerve. In a local service trade the demand does not expand much when price falls, so a general price cut mostly reallocates the same work at a lower margin. Within that, the shop that responds reliably takes wider damage than the shop that started it. Knowing why lets you answer a price move without answering it in kind.

Why a local service trade cannot win a price war

A price war can be positive-sum where a lower price brings in buyers who were not in the market at all. Residential and light commercial service mostly does not work that way. A failed water heater gets replaced this week at whatever the market charges. A no-heat call in January is not a purchase decision. A property manager's scheduled maintenance is a contractual obligation. That work arrives on its own schedule and is largely indifferent to a 10 percent price move.

There is a genuinely elastic slice and it is worth naming rather than pretending it away: deferral and scope. A customer sitting on a fifteen-year-old system decides this year or next year partly on price, and a customer choosing between a repair and a replacement moves on price. That slice is real. It is nowhere near large enough to cover what a general cut costs, which is the next section.

So the market total is close to fixed over a season, which means a lower price is not buying new work. It is buying a larger share of work that was going to happen anyway, at a worse rate per job, from competitors who can do the same thing to you.

The arithmetic of a ten percent cut

Take one job at a 38 percent gross margin, meaning the direct cost of delivering it is 62 percent of the price. Cut the price 10 percent and the cost does not move.

• Price goes from 1.00 to 0.90.
• Cost stays at 0.62.
• Margin per job goes from 0.38 to 0.28, a fall of 0.10.
• That 0.10 is 0.10 / 0.38 = 26 percent of the gross margin on that job.

To hold the same total gross margin you now need 0.38 / 0.28 = 1.357 times the jobs, so 36 percent more jobs than before the cut, at that starting margin, purely to stand still.

The starting margin matters enormously, and this is the sensitivity worth memorising:

Every figure in the middle column is stated against the original price so it can be compared with the left column, which is also stated against the original price. Expressed against the new, discounted price those same margins read 44, 31 and 22 percent, which is the number a spreadsheet will show you afterwards and is the reason a cut looks less damaging in hindsight than it was.

Read that against the previous section. The market is not going to produce 36 percent more replacements because two shops lowered their prices, and if it did, neither shop has the technicians to deliver them. The volume that would justify the cut does not exist.

Whoever holds more share loses more

Now the asymmetry in the title, and it has two independent causes. Be exact about what the effect actually tracks, because the title compresses it: the damage follows share, not the order of movement. A small challenger that matches a big incumbent's cut loses less by moving second, not more. The second mover loses more because the second mover is almost always the incumbent, which is who a challenger picks the fight with in the first place.

The first is share. Suppose the contested segment runs 100 jobs a quarter across the whole local market. You are the established shop with 45 of them. The new firm has 10. The rest are spread among others.

They cut 10 percent. Measured in the same unit for both sides - price-units of gross margin given up per quarter, inside the contested segment only - their cut costs them 0.10 x 10 = 1.0 unit. If you match, yours costs you 0.10 x 45 = 4.5 units. Same percentage, same segment, same quarter, and you give up 4.5 times what they do, entirely because you have 4.5 times the base to discount.

And the two sides are not buying the same thing with it. They are spending 1.0 unit to buy share, and if the cut works they end the quarter with more jobs than they started. You are spending 4.5 units to buy nothing: the best available outcome from a defensive match is that you keep the customers you already had, at a lower price.

The second is breadth. They chose the ground. Their cut applies to the segment they targeted, and a new entrant is usually concentrated in one or two job types. You will usually respond across your whole book, because that is administratively simpler, because your sales side hears price pressure as a general complaint, and because reacting in week two means nobody has segmented anything.

Put numbers on that. Say the contested segment is 55 percent of your revenue and your gross margin runs about 38 percent across segments - use your own segment margins here if you have them, this is a simplification. An across-the-board 10 percent cut costs you 26 percent of your total gross margin, as computed above. The cut bounded to the contested segment costs 0.10 x 0.55 = 0.055 of revenue, and 0.055 / 0.38 = 14 percent of your total gross margin.

Same cut, same competitor, same quarter: 26 percent of gross margin against 14, a little over half the damage, decided entirely by whether anyone drew a boundary around it before the price list changed.

The ratchet

The last piece of the asymmetry is that the two directions are not symmetric in time. A price falls in one decision and recovers across several, because customers form a reference price from what they were last quoted and revise it downward instantly and upward reluctantly.

Practically that means a cut you make in one afternoon takes several quoting cycles to walk back, each one a conversation with a customer who remembers the other number and has to be given a reason. Where the price was published - a website, a printed rate, an agreement renewal - it is worse, because the old figure is documented and the increase looks like a change in you rather than a change in the market.

The reference price also outlives the shop that set it, which is why a competitor going under does not restore your pricing. That mechanism is derived in full in a sibling card - see related: Competing Against a Shop That Does Not Know Its Own Costs.

What is not capitulation

Holding price is not the same as doing nothing, and the alternatives are the actual content of a response.

Compete on scope. A customer comparing two numbers takes the lower one. A customer comparing two documents that are visibly not the same job has to think. Itemise what is included, what is removed, what is left behind, what happens to the old equipment, who pulls the permit. The strongest version is not more inclusions, it is the same work stated so specifically that the other quote's silence becomes visible.

Compete on terms. A longer written labour warranty, a stated arrival window with a consequence attached, a fixed price rather than an estimate, a payment schedule that matches the customer's cash. Terms cost you something real, which is why they are credible, and they are much harder to copy in a week than a discount is.

Segment the response. If you decide to move on price, move on one job type, for one customer type, for a stated period. That is the difference between the 14 percent and the 26 percent above, and it is one decision.

A deliberate targeted response, and how to end it

Sometimes cutting is right. The conditions are narrow and worth stating, because the branch gets used far more often than it applies: a specific segment is genuinely being taken, the work has strategic value beyond its own margin (it feeds agreements, it keeps a crew busy in a trough, it holds an approved-vendor listing), and you have the margin to fund it for a defined period without borrowing.

If those hold, structure it so it can end:

• Make it an offer, not a rate. An offer has a name, a scope and an expiry. A rate change has none of those and becomes the new reference price the day it publishes.
• Bound it to the contested segment and say so internally, so nobody applies it to a job type nobody is competing for.
• Write the exit trigger down before you start. A date, plus a measurable condition: win rate back above a stated level, or the competitor's price moves, or a stated number of quarters elapses regardless.
• Name what you will do if the trigger fires and nothing has changed, which is usually to concede the segment rather than to extend. An offer that gets extended twice is a price cut with extra steps.

The one place a lower price is genuinely correct

There is a real exception and it is not a war at all, so it is worth separating out before someone uses it as cover.

While you are holding a crew through a trough, the wage is being paid whether the technician works or not. Split your direct cost into the part that is committed anyway and the part that is not - materials, fuel, subcontract labour, disposal - and only the second is a true cost of one more job. A job that covers that second part and contributes anything at all toward a wage you are paying regardless beats an idle hour.

Two guardrails, and without both this reasoning is how a shop ends up with a rate that never recovers its wages at all:

• It is dated and off-peak. The moment the extra job displaces a full-price job, the committed wage is no longer free and the whole argument collapses. That is a calendar condition, not a judgment call, so put dates on it.
• It does not reach customers who would have bought in season at full price. An offer that goes to your own base in November trains the base to wait for November.

Applied year-round rather than to a trough, this is the exact arithmetic that produces the shop that does not know its own costs - see related below. The Slow Season shelf owns how to fill a trough; what matters here is only what filling it does to your price. See related: The Slow Season Playbook.

How to tell you are in one, and how to tell you are not

Most shops that believe they are in a price war are in a win rate problem with a different cause, and the difference is diagnosable.

You are in a price war when the same competitor's number is named in the majority of your losses in a segment, when your own price has already moved once in response, and when their price moved again after yours did. That last one is the actual definition - a war requires a second move. One competitor pricing low is not a war, it is a competitor pricing low, and it has four possible causes with four different responses. See related: A New Competitor Undercuts You Sharply.

You are not in one when the losses are spread across different competitors, when customers name lead time or a poor quote experience rather than price, or when the drop is seasonal and matches the same quarter last year. Check the second and third of those before touching the price list, because a price cut aimed at a communication problem makes the shop worse at the price it now charges.

The failure mode to watch for is the quiet unilateral war: no competitor has moved at all, and your own quoting has drifted down because technicians and salespeople discount at the table to avoid a hard conversation. That produces every symptom above with no opponent. Measure the spread between your list price and your realised price by job type, and if the gap is widening while your list price is static, the war is internal.

References

• See related: Competing Against a Shop That Does Not Know Its Own Costs, A New Competitor Undercuts You Sharply
• See related: Why Pricing Against Your Competition Is a Trap, Competitive Positioning, Knowing Which Competitor You Are Actually Competing With
• U.S. Small Business Administration, pricing strategy and gross margin guidance for small business