I'm a procurement manager at a 220-person facility services company. For the last six years, I've managed a maintenance budget of about $480,000 a year, negotiated with 60+ vendors, and logged every invoice in our cost tracking system. I built a cost calculator after getting burned on hidden fees twice. That calculator is now the first thing I open when someone asks me to approve an order.
The question usually comes down to two options: Option A, fix what you have. Option B, replace it. The right answer doesn't come from the two quotes on the desk. It comes from comparing total cost of ownership across three dimensions: total cost over 24 months, time cost, and the risk that the same problem comes back. I used to skip the third one. Then a $340 repair became a two-season saga.
Dimension 1: Total Cost of Ownership, Not Just the Quote
Last spring, our commercial water heater started throwing the same error code. A contractor quoted $340 to replace the Honeywell hot water heater control and sensor. Seven months later, the same error returned. Another $340. When a third quote for the same fix came in at $380, we had already spent $680 and still had a unit that could fail at any moment. A new Honeywell hot water heater control with a three-year warranty, installed, came to $640. (Should mention: the $640 included labor, so there was no surprise setup fee.)
The obvious reading was 'repair is cheaper: $380 vs $640.' But the TCO reading was different. The repair path had failed twice in seven months. The upside of the third repair was saving $260 now. The risk was another failure during peak season, plus the same invoicing and work-order cycle. We replaced the control. I want to say the first repair was $340, but don't quote me on that exact invoice. The total after two failures came to $680, and that number was enough to make the replacement obvious.
The same logic applies to filters. We buy disposable MERV 8 filters for about $12 each and swap them monthly, which works out to $144 per unit per year. A washable K&N air filter cost about $60 in a standard one-inch slot. It took twenty minutes to clean every few months, and the breakeven landed around six months. I should add that the K&N air filter only worked in units where the MERV rating met code. In those units, the disposable filter became the expensive option, and the maintenance staff stopped burning time on monthly replacements.
Dimension 2: Time Is a Cost, Even When It's Small
Time is the line item that never appears on an invoice. A replacement might look simple on paper, but if it creates a service call, a wait for parts, and a follow-up visit, it eats budget sideways. I've had contractors quote a three-day lead time for a part that another supplier carried in stock down the street. The difference wasn't the price of the part; it was every hour that the space was down.
Take the office thermostat. The facility manager wanted to replace it because the building was freezing at 6 a.m. She had already picked out a new Honeywell thermostat. Before ordering, I asked if anyone had checked the schedule. Someone had set the overnight setpoint to 62°F while testing a sensor. The fix was learning how to reprogram a Honeywell thermostat, not buying a new one. It took four minutes, cost $0, and avoided a two-day service call. That was an uncomfortable conversation—the manager felt I was questioning her judgment. But it kept about $200 in the budget, and the thermostat still works.
Neck fans are a smaller version of the same mistake. We bought cheap $15 neck fans for crew members in an un-air-conditioned warehouse. Three out of eight broke during the first season. Every failure created a work order, a reorder, and a person working in the heat without the right tool. We then tested a $58 neck fan with a brushless motor and replaceable battery. It survived two full seasons. The $58 fan ended up cheaper than repeated $15 ones, and no one had to chase an invoice for it.
Dimension 3: Risk and the 'Again' Factor
The deciding dimension is almost always risk: how confident am I that this fix lasts? I look at failure history, not just part prices. If the same part has failed twice in seven months, the probability of a third failure is high enough that repair stops being sensible.
Someone asked me last month: why is my Samsung fridge not cooling but freezer works? That question sounds odd until you know how the system works. A frost-free fridge has a shared compressor for both compartments. If the freezer side stays cold but the fridge side goes warm, the likely culprits are an evaporator fan motor, a defrost heater, or a damper—not the compressor. A repair shop quoted $520 for a new motherboard. A second contractor found a seized evaporator fan and charged $210. The $210 repair won because the failure was a simple part and the fridge was still young. If the same Samsung fridge were twelve years old and the diagnosis pointed to a sealed-system leak, I would replace it, because that repair can cost more than half the price of a new unit and has a real chance of failing again.
Actually, I should add that the second contractor checked the frost pattern before ordering any parts. That little step is what made the repair quote reasonable instead of a shot in the dark. It is also why I ask vendors to explain the likely cause before I approve a repair—not just give me a number.
So When Do I Hit 'Replace'?
After running this comparison for years, I have a few simple rules. They aren't universal, and I break them when the circumstances are unusual, but they keep me from defaulting to whichever quote is cheaper at that moment.
- Repair if the equipment is under five years old, the failed part is common, and this is the first failure.
- Replace if the same fix has needed to be done twice in 12 months.
- Replace if the repair quote is more than 50% of the replacement cost and the equipment is past its expected service life.
- Before both options, check the settings. Sometimes the cheapest fix is a four-minute reprogram.
For HVAC equipment, I also use Energy Star's rule of thumb as a sanity check: once a unit passes 15 years old, replacement should be on the table even if the repair quote looks small. I don't treat that as gospel. I treat it as a nudge to run the full TCO model instead of just looking at the invoice.
The point is not to avoid repairs. Some repairs are the right answer, like a $210 fan motor on a young Samsung fridge. The point is to compare the right things before defaulting to the cheaper quote. A repair can be cheap today and expensive next month. Replacement can be expensive today and cheap over the next three years. Running the numbers on total cost, time, and risk is the only way I've found to tell the difference.
