A machine that's booked on every job for months can still be losing money, and a machine that sits idle three days a week can still be the right one to own. The percentage everyone calls "utilisation" doesn't distinguish between the two, because two genuinely different calculations share the same name.
Key takeaways
Hours-based utilisation is straightforward: hours the machine actually worked, divided by hours it was available to work. It answers “how busy was this machine,” and it's the number most equipment-tracking software reports by default.
Cost-recovery utilisation answers a different question: is the revenue or internal job-cost credit this machine generated enough to cover what it costs to own, regardless of how many hours that took. A machine can score high on the first measure and still fail the second, if the rate it's being charged out at doesn't actually cover its fixed ownership cost — financing, insurance, standing maintenance — whether it worked one day that period or twenty.
Procore’s Equipment tool lets a company “track and manage” equipment, “track time for… equipment in timesheets,” and “see time entered for equipment in your budget” — which supplies the hours-worked side of the calculation directly, and feeds into the budget under cost type “E” (Equipment) alongside labour, materials and the other default categories.
What the tool doesn't do on its own is decide whether those hours, at whatever internal or external rate they're billed, actually cover the machine's fixed ownership cost. That comparison — hours logged against the fixed cost the machine is carrying — has to be run separately, using the timesheet data as an input.
The regulations carve out a specific CCA category — “contractor's movable equipment, including portable camp buildings, acquired for use in a construction business” — which is the CCA class most owned construction equipment falls into, setting the pace at which its fixed cost is written off for tax purposes. That depreciation schedule is one input into the machine's true annual fixed cost, alongside financing and insurance, and the annual fixed cost divided by hours available is the baseline cost-per-hour a machine has to clear before it's earning its keep.
The buy-or-rent break-even calculation covers working out that fixed annual cost and dividing it by the daily (or hourly) rental rate for the same class of machine to get a break-even utilisation — the same logic applies whether the comparison is against renting the equivalent machine or against the internal rate the company charges its own jobs for using owned equipment.
A fleet assembled piece by piece, each purchase justified against the utilisation expected at the time, can end up carrying machines that individually made sense but collectively no longer clear their fixed cost once the work that justified them has moved on. Re-running the cost-recovery calculation against actual logged hours — not the hours originally forecast — on a periodic basis is what catches that drift before it's been quietly costing money for a year.
There’s a tax-side consequence to actually retiring a machine that's stopped clearing its keep, too: CRA's capital cost allowance guidance confirms that when a piece of equipment is sold or scrapped and there's no property left in its CCA class, any remaining undepreciated capital cost becomes deductible in full as a terminal loss the same year, rather than continuing to depreciate an asset that's no longer earning anything — one more reason a fleet review has a real payoff beyond the utilisation number itself.
Automated hour-meter and location data — whether from a telematics system or from equipment-tool timesheets like Procore's — removes the guesswork from the hours-worked side of the calculation, replacing an estimate of how much a machine was used with an actual logged number. That's a real improvement over manual logs, which tend to under-report idle time nobody bothers writing down.
What automated data collection doesn't do on its own is decide whether those hours cleared the machine's fixed cost — that comparison still has to be run against the specific machine's ownership cost and the rate it's actually being charged out at. the preventive-maintenance article covers the operational side of building that data collection into a mixed fleet's routine; the utilisation math still has to be run separately from whatever system is generating the hours.
A worked example
A skid steer is available 220 working days a year and logs 140 of them — a 64% hours-based utilisation rate, which looks reasonably healthy on its face.
Its fixed annual ownership cost (financing, insurance, standing maintenance) runs $24,200.00. At an internal charge-out rate of $210.00 a day credited against jobs for the 140 days it worked, it generated $29,400.00 in cost recovery — comfortably above the $24,200.00 fixed cost, so this machine clears its cost-recovery bar even though its hours-based number (64%) isn't especially high. A machine doesn't need to work every available day to be worth owning; it needs to clear its fixed cost, which is a different and lower bar in this case.
Contrast a second skid steer on the same yard, logging only 60 days a year at the same $210.00 rate: $12,600.00 in cost recovery against the same $24,200.00 fixed cost — a $11,600.00 shortfall. Both machines are “utilised” in the loose sense that they worked some days. Only the calculation tied to fixed cost tells you the second one is losing money.
No — a machine can be booked on every available day and still not be earning its keep if the rate charged against those jobs doesn't cover its fixed ownership cost. The goal is clearing the cost-recovery bar, not maximising hours worked for its own sake.
It depends on the specific machine's fixed ownership cost and the going rental rate for the same class — the buy-or-rent article covers working out that break-even in days per year, which varies by machine, not a single rate that applies across a fleet.
Procore’s Equipment tool tracks the hours-worked side through timesheets and equipment records feeding the project budget, but it doesn't run the cost-recovery comparison against a machine's fixed ownership cost on its own — that calculation has to be assembled separately from the timesheet data it produces.
At least annually for every owned machine, and sooner for anything acquired for a specific job that's now finished — a machine bought to justify one contract's utilisation forecast is exactly the kind of asset that needs re-checking once that contract ends, not left running on the original assumption indefinitely.
A 30-minute call is enough to tell you whether AI pays for itself here.
