Road contractor: machine-day rate by season, not by calendar
“I counted a machine-day the way a bookkeeper counts: take the yearly cost of the machines and divide it by the working days. It came out at 24 500 ₴. The right figure was 42 000 ₴, and for eight years I handed the customers my own margin in the gap between those two.”
Roman Tsymbaliuk builds roads in Ivano-Frankivsk. Sixteen years in road works, eight of them as the owner. A paver, two rollers, a milling machine, three tippers. The crew is eighteen people. Jobs for the district and the city, all through tenders.
The season closes with 1 476 000 ₴ of profit on revenue of 58 800 000 ₴. Two and a half per cent. Roman thought that was the price of competition. It turned out to be the price of one division.
Where the money is in road works
Almost nowhere, and for this trade that is normal.
| Item | For the season |
|---|---|
| Revenue | 58 800 000 ₴ |
| Buying the mix | 40 320 000 ₴ |
| Fuel and emulsion | 4 320 000 ₴ |
| Crew | 4 410 000 ₴ |
| Machines | 6 174 000 ₴ |
| Administration | 2 100 000 ₴ |
| Profit | 1 476 000 ₴ (2.51%) |
Why the margin here is always thin
In road works the profit never lies in the big lines. It lies in how each of the small ones is counted.
The mix is 68.6% of revenue. You do not control its price: the plant publishes a price list and there is almost nothing there to bargain over. All you control is the other thirty per cent.
When two thirds of the estimate is set from outside, a mistake in the remaining third eats not a share of revenue but a share of profit. And profit here is two and a half per cent.
That is why in road works there is no such thing as a small mistake in a rate.
The machines do not know about winter
A paver costs money every single day.
Over a year that comes to 6 174 000 ₴, and it is made up of things not one of which depends on how many jobs you have:
- lease payments on the paver, the rollers and the milling machine
- insurance on the machines and third-party liability
- storage on a guarded yard
- depreciation and the wages of two mechanics through the off-season
That sum does not ask whether you are working. It is the same in July as in January.
And asphalt can be laid for seven months. From April to October, while the temperature holds. In November you can no longer lay it, and the machines go onto the yard.
Anyone who has once watched hot mix tipped out into the cold knows this: it sets before it can be levelled, and the surface then comes away in sheets with the very first spring. So the season closes by the thermometer, not by the owner's wishes.
“I always knew that we do not work in the winter. I just never once thought that the lease does not know it.”
One division, two rates
There are twenty-one working days in a month.
- twelve months make 252 working days
- the seven months of the season make 147 working days
- 6 174 000 ₴ divided by 252 gives 24 500 ₴
- 6 174 000 ₴ divided by 147 gives 42 000 ₴
The difference is 17 500 ₴ on every machine-day.
On one job that goes unnoticed. Over a season with many jobs it is already the whole sum the owner wonders about every year, when there is no profit although the machines are loaded to the brim.
Why it works out exactly like that
The whole difference rests on one thing: twelve against seven. The season rate is 1.714 times the calendar rate, and that is exactly the ratio of the months. There is no other mathematics here.
And it is not twice. The temptation to say “twice as expensive” is strong, but the figure is a different one, and in an estimate it behaves differently.
What the percentage is taken from
This is easy to get lost in, and Roman got lost in it too.
The shortfall of 17 500 ₴ is 41.7% of the correct rate of 42 000 ₴. The same shortfall is 71.4% of the wrong rate of 24 500 ₴.
One and the same mistake has two sizes, and that depends only on what you compare it with.
“My estimator told me: we were out by seventy per cent. The same week the customer heard from me that we were under by forty-two. Both of them were right, and for two days I could not understand how that was possible.”
When you are talking to the customer, the base is the correct rate. When you are showing the estimator how far he was out, the base is the wrong one. They must not be mixed up.
On asphalt you cannot see it
And that is exactly why the mistake lived for eight years.
In an ordinary laying job the mix takes almost seventy per cent. The machines are a small part of that estimate, and a shortfall of 17 500 ₴ a shift drowns in the mark-up on the material. The job closes in the black and nobody goes checking.
The mistake becomes visible where there is no mix in the order at all.
A road contractor has few such orders, and they always look secondary. That is precisely why nobody checks them separately: they bring in little revenue, and it seems there is nothing there to get wrong.
An order with no mix
Milling off the old surface. A purely mechanised process, almost no material in it.
Twelve thousand square metres at 86 ₴ a metre — revenue of 1 032 000 ₴. Fourteen machine-days are needed.
| Item | In the estimate | In reality |
|---|---|---|
| Revenue | 1 032 000 ₴ | 1 032 000 ₴ |
| Machines | 343 000 ₴ | 588 000 ₴ |
| Crew | 176 400 ₴ | 176 400 ₴ |
| Fuel and picks | 137 200 ₴ | 137 200 ₴ |
| Profit | 375 400 ₴ | 130 400 ₴ |
What happened here
One line changed. Fourteen machine-days at 24 500 ₴ is 343 000 ₴. The same fourteen at 42 000 ₴ is 588 000 ₴.
The margin fell from 36.4% to 12.6%. Of the estimated profit, 34.7% is left — a little more than a third.
Not one other figure in the estimate changed. The same area, the same price per metre, the same crew, the same fuel. The mistake was in one multiplier.
And this has to be said plainly.
The 130 400 ₴ is a plus. Milling did not eat money, it brought money in. It simply brought in 2.88 times less than the estimate showed.
“If it had come out negative, I would have seen it in the very first year. But it came out positive, only small. And every time I thought I had simply charged too little per metre.”
For eight years Roman was convinced that milling was a weak order, and he tried to take less of it. In fact what was weak was the estimate.
This is the most expensive kind of mistake: it does not spoil the figures, it spoils the decisions. For years Roman turned down orders that were perfectly sound and took on the ones where the shortfall simply hid inside the material.
What was tried before
The first attempts were about price, not about the rate.
Roman put up the price per metre for milling and lost part of his tenders because of it. Then he tried to squeeze the crew and the fuel, but all that could be wrung out of there was percentages, while the machine line held the whole difference between 343 000 ₴ and 588 000 ₴.
Not one of those moves touched the line the mistake was in.
There was a third attempt as well, the most logical of them all: sell the milling machine. If the order is weak, why keep the machine that does it. Roman worked it out and saw that without the milling machine he also loses the preparatory work on his own jobs, and subcontracting that comes out more expensive.
Why the calendar rate is attractive
Because it wins you the tender.
“We used to win tenders, and I took that for my strong side. Not one competitor could understand how we gave such a price. Now I think they were counting exactly the same way, they just had fewer machines.”
A lower rate means a lower estimate, which means a better chance. Roman won jobs he would never have won with the right figure, and he sincerely took that for an advantage. The machines are standing there anyway, let them earn something.
Where the substitution happens
The machines are not “standing there anyway”. They cost 6 174 000 ₴ regardless of anything, and somebody has to cover that money. If not the customer, then the owner.
The phrase “let them earn something” only makes sense when the machine really would otherwise stand idle. But in the season it does not stand idle: the number of machine-days is limited, and every one given to a cheap job is a machine-day that will not exist for a dearer one.
A job won at the wrong rate adds no profit. It adds turnover and takes away machine-days that could have gone to a job at the right price.
The move that works
The correct rate changes nothing by itself. It only shows the size of the hole.
There are two ways to close it. The first is to raise prices so that 42 000 ₴ stands in the estimate. That is a share of tenders you will no longer win.
The second is to bring the rate itself down. Not through costs, but through the denominator.
Why the second way is easier
The first way needs the market to agree with you, because every hryvnia in the estimate has to pass a tender committee and a competitor who counted just as wrongly as you did eight years ago. The second needs the calendar to agree with you.
The calendar bargains more willingly than the customer.
Stretching the window by a month
Asphalt is not laid in the cold. But road works are not only asphalt.
Pothole repair with cold mix can be done in November too. It is not the order a road contractor dreams of, and the margin there is modest. What it does give you is twenty-one working days.
- the months of work become eight, and the working days 168
- 6 174 000 ₴ divided by 168 gives 36 750 ₴
- the rate fell by 5 250 ₴, that is by 12.5%
- and it is still higher than the calendar 24 500 ₴
| Work window | Working days | Machine-day rate |
|---|---|---|
| Calendar year | 252 | 24 500 ₴ |
| Asphalt season | 147 | 42 000 ₴ |
| Season plus November | 168 | 36 750 ₴ |
The last line here is the most important one.
Stretching the window does not make the calendar rate correct. It only narrows the gap a little. There are twenty-one more working days, and the same yearly sum spread itself thinner.
The calendar is always wrong
Even with the stretched window the correct rate is 36 750 ₴, not 24 500 ₴.
Dividing by the calendar would be right only if the machines worked all twelve months with no idle time whatsoever. In road works that does not happen in principle.
A machine-day rate is divided by the days the machine works, not by the days the calendar exists.
What November brought
The first year came out cautious.
Twenty-one days of pothole repair.
Roman took no big jobs and promised customers no deadlines he could not hold in the cold. He took what others take reluctantly: pothole repair in courtyards and on access roads.
Revenue of 1 806 000 ₴, costs of 1 092 000 ₴. The effect is 714 000 ₴.
That is 48.4% of the profit for the year. Almost half, though half it did not reach.
Why so much
Because on those days the machines add not one hryvnia of cost. The 6 174 000 ₴ has already been paid for the whole year; in November the machines simply stopped standing still. All that had to be covered was the cold mix, the crew and the fuel.
“We worked that first November without a single new machine and without a single new man. We just stopped putting the machines on the yard a month earlier.”
Where else the same mistake sits
The machine-day rate is not the only place where dividing by the calendar spoils an estimate.
The crew's wages are counted the same way, when the yearly payroll is spread over twelve months although the crew works seven. Administration is counted the same way, and that is exactly why in the off-season it always looks unbearably expensive. The logic of the mistake is one and the same: the cost exists all year, it earns only in the season, and it has to be divided by the days when it earns.
The difference is that with the crew this is visible at once, because the wages are paid in live money every month and the owner watches them run into zero over the winter. The machines are paid for by lease, and a lease does not feel like a cost — it feels like an obligation.
Where the second base is here
The same figure again has two sizes.
The 714 000 ₴ is 48.4% of the old profit of 1 476 000 ₴. In the new profit of 2 190 000 ₴ that same sum is 32.6%.
Both figures are true. The first answers the question “how much did we grow”, the second answers “what share of the new result came out of November”. Mixing them up is just as dangerous as mixing up the two rates.
When Roman told other road contractors about November, he quoted the 48.4%, and every time it sounded like an invention. He only had to add the words “of the previous season's profit” and the disbelief went away.
What to do with the old estimates
Recalculate not all of them, only the ones where the machines weigh more than the material.
- milling without supplying the mix
- grading and base preparation as a separate order
- rolling somebody else's mix
- any work where the machine is yours and the material is the customer's
“I recalculated four types of estimate in one evening. Three of them we have charged more for ever since, and one we stopped taking altogether.”
How to recalculate quickly
Take the estimate, find the machines line, multiply the number of machine-days by 42 000 ₴ instead of 24 500 ₴ and look at what is left of the profit. If less than half is left, you have been selling that type of order cheaper than you thought for years.
In ordinary laying the difference smears itself into the mark-up on the tonne. In these four cases it eats two thirds of the profit, and no other line makes up for it.
The recalculation needs neither software nor a consultant. Two figures are enough: the yearly cost of the machines and the number of days those machines actually leave the yard. The second figure is in every works log, it is just that nobody has ever added it up for the year.
The money does not vanish — you simply do not see it
Roads look like a business about asphalt and tenders. In fact they are about how many days a year your machine leaves the yard.
Roman did not have a problem with competition. He had a machine-day rate counted on the days when the machines are not working.
All that was needed was to divide the same sum by a different number. The machines did not get cheaper, the lease was not rewritten, not a single machine was sold or bought.
The difference between 24 500 ₴ and 42 000 ₴ is 17 500 ₴ a shift, and for eight seasons it went to the customers as a gift. One month of November brought back 714 000 ₴, because for the first time it divided the same sum by a larger number of days.
How to see this in your own estimates
All the arithmetic above rests on one row the estimate does not have: how many days a year the machines actually leave the yard. While fixed equipment costs are divided by calendar days, neither the true machine-day rate nor the jobs that were given away for years is visible. When the same data is collected automatically, the rate is visible for every site separately — which is exactly why owners of small and medium businesses use Finmap: to see the financial picture rather than gather it from fragments.
Read also
Topic foundation: Management Accounting Explained: What It Is and Why Owners Need It
Questions that come up right away
Divide the yearly fixed cost of the machines by the days the machines really work, not by the calendar working days. Roman's 6 174 000 ₴ over the 147 days of the season gives 42 000 ₴, and over 252 calendar days only 24 500 ₴. The difference is 17 500 ₴ on every shift, and it equals the ratio of twelve months to seven.
Because that is how the bookkeeping counts: the costs of the year over the working days of the year. It is correct for the books and wrong for an estimate. The calendar rate would be right only if the machines worked all twelve months with no idle time, and in road works that never happens.
Because buying the mix takes 68.6% of revenue, and the machines are a small part of that estimate. A shortfall of 17 500 ₴ a shift drowns in the mark-up on the material, the job closes in the black, and nobody goes checking. It becomes visible where there is no material.
On 12 000 m² at 86 ₴ the revenue is 1 032 000 ₴, and fourteen machine-days are needed. In the estimate the machines cost 343 000 ₴, in reality 588 000 ₴. Profit falls from 375 400 ₴ to 130 400 ₴, so 34.7% is left — a little more than a third, and the margin goes from 36.4% to 12.6%.
No, it is profitable: the 130 400 ₴ is a plus. It is simply 2.88 times poorer than the estimate showed. That is exactly why the mistake lives for years: a loss is visible at once, while a small plus gets blamed on competition and on a low price per metre.
The working days become 168 instead of 147, and the machine-day rate falls from 42 000 ₴ to 36 750 ₴, that is by 12.5%. Twenty-one days of pothole repair with cold mix brought revenue of 1 806 000 ₴ against costs of 1 092 000 ₴, so 714 000 ₴ of effect.
Because on those days the machines add no cost: the 6 174 000 ₴ has already been paid for the whole year. All that has to be covered is the cold mix, the crew and the fuel. In the old profit of 1 476 000 ₴ that 714 000 ₴ is 48.4%, and in the new profit of 2 190 000 ₴ it is already 32.6% — two different bases.
The ones where the machines weigh more than the material: milling without supplying the mix, base preparation as a separate order, rolling somebody else's mix and any work where the machine is yours and the material is the customer's. In ordinary laying the difference smears itself out; in these cases it takes two thirds of the profit.
