Labour Cost Rising Faster Than You Can Reprice

Why this matters

This is the shock that closes good shops, and it closes them quietly. Nothing dramatic happens: the phone still rings, the trucks still roll, the work is still good. Wages move, prices do not move with them, and eighteen months later a shop that has never had a bad month has no reserve left and cannot say where it went. The reason it is hard to see is that the loss is spread across every invoice rather than concentrated in an event. The reason it is fixable is that the whole thing is arithmetic, and the arithmetic is small enough to do on a Sunday afternoon.

Two clocks, running at different speeds

Your wage clock is set by the labour market and it is not optional. When a competing shop, a warehouse or a utility raises what it pays, you either match it within a pay period or two or you start losing people. You do not get to decide when this happens, and you cannot phase it. It is forced, it is fast, and it applies to your whole crew at once.

Your price clock is set by you and your customers, and it is slow. You decide to raise rates, you update the price book, the change reaches new quotes, and then it has to work through everything you have already sold. Every job in the backlog, every agreement mid-term, and every fixed-price contract carries the old price against the new wage.

The gap between those two clocks is the whole subject. It is not a vague pressure, it is a measurable number of months during which you are paying tomorrow's wages on yesterday's prices.

What the squeeze actually costs

Work in percentages of revenue so the answer travels. Take a shop with direct labour, fully burdened, at 32 percent of revenue - burden meaning wages plus payroll taxes, comp, benefits and the rest, which Fully Burdened Labor Rate Calculation derives and this card does not. Materials are 25 percent, other direct costs 8 percent, overhead 25 percent, leaving net profit at 10 percent of revenue. Those add: 32 plus 25 plus 8 plus 25 plus 10 equals 100.

The local market moves entry pay up 8 percent. After the band adjusts (below), the shop's burdened labour cost rises 6 percent.

Labour was 32 points of revenue, so it becomes 32 x 1.06, which is 33.92 points, and the increase is 1.92 points of revenue. Nothing else changed, so the whole 1.92 comes out of the 10 points of net profit, leaving 8.08 percent net.

Say that in the unit that matters: a 6 percent wage move took 19.2 percent of the shop's net profit - 1.92 of the 10 points it had - on a year in which revenue did not fall by a cent and nobody did anything wrong.

The price move that restores it is smaller than owners expect

Owners routinely assume a 6 percent labour rise needs a 6 percent price rise, decide that is unsellable, and then do nothing at all. It does not.

Total cost after the wage move, still measured against the old revenue base of 100, is 33.92 labour plus 25 materials plus 8 other direct plus 25 overhead, which is 91.92. To hold net at 10 percent of the new revenue R, you need R minus 91.92 to equal 0.10 R, so 0.90 R equals 91.92 and R is 102.13. Check it: revenue 102.13 less cost 91.92 leaves net 10.21, and 10.21 over 102.13 is 10.0 percent.

A 2.13 percent price rise restores the margin percentage. Restoring the net amount rather than the percentage takes slightly less, 1.92 percent, since you only need to cover the 1.92 points. Either way the answer is about a fiftieth, not a sixteenth, and it is a conversation that can actually be had with a customer.

The thing that makes it expensive is not the size of the increase. It is how long the increase takes to reach the work.

The three pools that cannot be repriced today

Sort revenue by how quickly a price change reaches it. The months below are that pool's average time at the old price, counting your own decision and rollout lag as well as the contractual one.

The shares sum to 100 percent and the weighted months total 4.65, which is 38.75 percent of a year (4.65 over 12). So even a shop that reprices fully and promptly eats the compression across more than a third of the first year: 1.92 points x 0.3875 is 0.74 points of revenue, which against a 10-point net is 7.4 percent of one year's net profit, spent entirely on the lag.

That 7.4 percent is the price of being slow once. The 19.2 percent from the previous section is the price of never repricing at all, every year, compounding. The two numbers answer different questions and it is worth keeping them apart.

Run your own version of that table first, because the mix decides the answer. A shop that is 90 percent spot work has a weighted lag near three months and barely notices. A shop with half its revenue in the table's fixed-price commercial row and the rest in spot work has a weighted lag near six months (4.5 plus 1.25), nearly half the year at the old price, and is in a genuinely different business.

Why the bottom of the band moves the whole band

The gap between the 8 percent move in entry pay and the 6 percent move in the shop's blended labour cost is not rounding. It is what happens when you try to raise only the bottom.

Suppose a second-year technician was earning 15 percent above entry. Entry goes to 1.08 of its old level, the second-year rate stays at 1.15, and the differential collapses from 15 percent to 1.15 over 1.08, which is 6.5 percent. That technician now does considerably more skilled work for a margin over a new hire that has more than halved, and they will notice within one pay cycle because the new hire tells them what they are making.

So the realistic version is that you move the band, not the entry rate, and the blended increase lands somewhere between the two. Budgeting the entry-rate move alone and being surprised by the rest in month three is the ordinary way this gets mishandled.

The levers, and what each can actually carry

Repricing, on a trigger rather than a calendar. The largest lever and the one most often deferred. See the cadence below.

An escalation clause in every recurring agreement. This is what shortens the 6.0-month and 9.0-month rows in the table. The shape customers accept is a stated annual adjustment tied to a named published index rather than to your judgement, with a cap, and written notice a set number of days before the anniversary. A published compensation index is legible and neutral in a way that "we reserve the right to adjust pricing" is not. Drafting it is the Contracts shelf's job, not this card's, and a clause in a multi-year commercial agreement is worth your own attorney's time before it goes into the template you will use a hundred times.

Productivity, with an honest limit. To offset 1.92 points through productivity alone you need labour hours per unit of revenue to fall by 1.92 over 33.92, which is 5.66 percent - check it: 33.92 x (1 minus 0.0566) is 32.0, back where you started. A 5.7 percent cut in labour hours is real work: drive time, callback rate, parts on the truck, second trips. It is achievable over a year or two and it is not achievable in a quarter, and telling a crew to work 6 percent faster is not a productivity plan.

Mix, with a warning attached. Shifting toward work where labour is a smaller share of the ticket lowers your exposure to the next wage move. The warning is that this lever runs backwards in a downturn: the deferrable, material-heavy replacement work is exactly what customers stop buying, and the repair work that replaces it is labour-heavy, so a soft market raises your labour share at the same time a tight labour market raises your labour cost. Those two shocks compound, which is why a wage squeeze during a slowdown feels so much worse than either alone. See related: Deferrable Versus Non-Deferrable Work and What a Downturn Touches.

The cadence, and the trigger that replaces the calendar

An annual rate review adds an average of six months of detection lag on top of the 4.65 months of contractual lag, and in a fast labour market that is most of a year at the wrong price.

A workable default, which you should tune: recompute the burdened labour rate quarterly, and change billing rates whenever the burdened rate has moved more than 3 percent since the last billing-rate change, regardless of where you are in the calendar. Quarterly is often enough to catch a move inside one billing cycle and rare enough that the office will actually do it. The 3 percent trigger keeps you from renegotiating over noise while still catching a 6 percent move on the first review after it lands.

Two things have to happen the same day the rate changes, or the change does not reach the money: the quote validity window on outstanding proposals gets checked against the new rate, and the agreement renewal list for the next two quarters gets the new rate applied at renewal. See related: Quoting While Prices Are Moving, which owns validity mechanics.

Running your own numbers

Print each figure beside the test it is being checked against:

• The cost shares sum to 100 and net is stated as a share of revenue. Above: 32 plus 25 plus 8 plus 25 plus 10 equals 100, and net is 10 percent of revenue, not 10 percent of cost. A margin quoted on the wrong base will make every downstream figure wrong in the flattering direction.
• The compression is quoted with its base in the same breath. 1.92 points of revenue, which is 19.2 percent of a 10-point net. Reporting it as "under 2 percent" is true and useless.
• The pool shares sum to 100 and the weighted lag is in months. 60 plus 15 plus 25 equals 100; 1.50 plus 0.90 plus 2.25 equals 4.65 months. If your pools do not sum, the missing share is sitting in whichever row you estimated.
• The productivity offset is stated on the post-increase labour figure. 1.92 over 33.92 is 5.66 percent. Computing it on the pre-increase 32 gives 6.0 percent, which is the wrong denominator and overstates what you need.
• The trigger has a number and a cadence. Quarterly recompute, 3 percent move. "We review rates regularly" is not a trigger.

References

• U.S. Bureau of Labor Statistics, Employment Cost Index (total compensation, private industry) and Occupational Employment and Wage Statistics, for the published index an escalation clause can name and for local wage levels
• See related: Fully Burdened Labor Rate Calculation (owns the burdened-rate calculation), Labor Burden: The Real Cost of an Employee
• See related: Quoting While Prices Are Moving (owns quote validity and escalation mechanics), Deferrable Versus Non-Deferrable Work and What a Downturn Touches
• See related: the Contracts shelf, for drafting an escalation clause into a recurring agreement