Most facilities managers know what they spend on maintenance each month. They can pull the number from the GL — parts, labor, contractors. What very few can tell you is how much each individual piece of equipment costs to maintain. That gap is where the real money is hiding.
Why total maintenance spend is the wrong number
Total maintenance spend tells you how much you're spending. It doesn't tell you where it's going, whether it's justified, or which assets are quietly becoming money pits. Two facilities spending identical amounts on maintenance can have completely different asset performance profiles — one with a few problem assets driving the majority of costs, another with spend distributed evenly across a well-maintained fleet.
You can't make good asset replacement decisions with total spend. You can with cost-per-asset. And the difference between a data-driven "replace it" decision and a gut-feel one is usually five figures.
The full cost equation
Most teams that attempt to track per-asset costs only count parts. That's not the full picture. The true maintenance cost per asset over a period includes:
- Parts and materials — everything ordered specifically for that asset, including consumables used during service
- Internal labor — technician time spent on the asset, multiplied by your loaded labor rate (salary + benefits + overhead)
- Contractor costs — any external invoices directly attributed to that piece of equipment
- Downtime cost — the production or operational value lost while the asset was down. This is the one teams consistently undercount, because it requires knowing your downtime cost per hour
A boiler that costs $2,000 per year in parts might cost $14,000 per year once you account for labor and downtime. That changes the conversation about whether to repair or replace it.
The replacement threshold calculation
Once you have cost-per-asset data, the most valuable use of it is evaluating whether an asset should be replaced. The rule of thumb used by most maintenance engineers: if annual maintenance cost exceeds 40–60% of the asset's replacement value, replacement is likely more economical than continued repair.
For older assets, add a second factor: reliability trend. An asset that's costing 30% of replacement value annually but has required three unplanned emergency repairs in the last six months is a different risk profile than one with the same spend driven by scheduled PMs. Frequency and severity of breakdowns matter as much as total cost.
What you need to make this work
Calculating per-asset costs manually — even for a small asset register — is a significant effort. You need every work order tied to a specific asset with parts costs and labor hours captured. You need technician time tracked at the task level, not just submitted as a daily total. And you need that data going back far enough to spot trends.
That's the gap that a CMMS is designed to fill. When every work order is created against a specific asset, and parts usage is logged at order close, the system builds the per-asset cost history automatically. Reporting becomes a filter, not a manual calculation exercise.
The question most facilities ask after seeing their per-asset cost data for the first time isn't "what do we do now?" — they already know. The question is "how did we not see this sooner?" The answer is usually: they didn't have a system that made it visible.
Getting started without historical data
If you're starting from scratch — no CMMS, no structured records — you don't need to wait for years of history before making better decisions. Start now with what you know:
- Set up your asset register and tag every piece of equipment
- Start logging every work order against a specific asset from today forward
- Capture parts costs and labor hours at close-out, every time
- After 90 days, you'll have enough data to start seeing which assets are outliers
Three months of clean data beats three years of guessing. The cost-per-asset view will surface problem assets faster than you expect — and some of them will surprise you. Start tracking your asset costs in Shiftlyio →
