Of the four cost lines that decide whether a restaurant P&L is fixable in a quarter, repair and maintenance is the one operators most often ignore. Labor is loud. Food cost is measured. Comps and voids at least get talked about. R&M sits in a corner of the P&L, drifting up 20 basis points a quarter, and nobody notices until the annual review shows the line has grown 60 percent over two years.
That is R&M creep. And it is the most preventable of the four cost problems, once you understand the mechanism. The problem is not the equipment. It is the reactive workflow around the equipment.
What R&M creep looks like on the P&L
Healthy R&M runs 1 to 2 percent of sales, and the line is flat or gently declining over quarters as newer equipment ages into peak reliability. R&M creep looks like this:
Fig. 1 · R&M creep is invisible one quarter at a time.
The chart is real. This is a location in a group I audited in 2024. R&M ran 1.3 percent of sales in Q1 2022 and 3.9 percent in Q4 2023, an eight-quarter climb of 2.6 points. On a $6M unit that is $156,000 of annual R&M spend when the healthy number should have been $120,000. The gap of $36,000 is the invisible cost of a broken maintenance workflow.
R&M creep is not a story about equipment. It is a story about the workflow around equipment. The same walk-in fails eight times in a quarter because nobody wrote down that it failed the first seven.
The mechanism, in detail
Here is exactly how R&M creep happens. Same story, every restaurant.
Tuesday, 10:47am. The walk-in door seal is compromised, condensation is dripping onto product. The kitchen manager pulls out her phone, calls the handyman she has used for two years. He comes in at 2pm, tapes and re-clips the seal, takes $180 in petty cash, hands the kitchen manager a paper receipt.
The kitchen manager puts the paper receipt in the petty cash envelope. She does not photograph it. She does not enter anything into a log. Petty cash reconciles at end of week, the $180 posts to R&M expense, and life moves on.
Three weeks later, same walk-in, same seal, same failure. Same handyman, different day, another $180. Six weeks later, again. In one quarter the walk-in seal has cost $720 in reactive repairs. The actual seal, replaced properly with the correct part and torque on the mounting screws, is a $340 fix that takes two hours and does not fail again for two years.
You have paid $720 to defer a $340 fix. And because none of the repairs got logged, when the R&M line goes hot at quarter end, nobody in the building can tell you why. All they can tell you is "equipment is old." Which is not the truth. The truth is "we have paid four times to patch the same failure and never fixed it properly."
The 30-day log that makes the fix possible
The first move is a log. Any shared document. A spreadsheet. A simple app. The tool matters less than the discipline. Required fields:
- Date and location so patterns by location are visible.
- Equipment and nameplate photo so the specific unit is identified, not just "the fridge in the back."
- Description of failure in plain language: "walk-in door seal dripping condensation onto lower rack."
- Temporary fix applied: "handyman re-taped and re-clipped seal, replaced two mounting screws."
- Invoice amount and vendor for cost tracking.
- Photo of the receipt. Two seconds. Prevents the "we lost the receipt" pattern that hides half of R&M spend.
The rule that makes the log actually get filled in: no petty cash without a log entry. The general manager checks the log before approving petty cash reconciliation at end of week. If a $180 repair does not have a log entry, the manager who authorized it fills it in before the petty cash gets replaced.
Give it 30 days. By day 30 you will have between 15 and 40 entries per location. Sort them by equipment. Three or four items will account for 60 to 80 percent of the total R&M spend. Those are your root-cause targets.
What the log reveals
On the first group I ran this on, the 30-day log across five locations surfaced this pattern:
Fig. 2 · Pareto shows up cleanly in R&M data.
Three items accounted for 87 percent of the R&M spend that month. Walk-in refrigeration, ice machines, dishwashers. And when we drilled into each one, the story was the same: repeated patches on the same underlying failure that no one had ever fixed root cause. Walk-in seals, ice machine water filters, dishwasher door gaskets. Small components, large cumulative bills.
Root cause versus patch
The difference between root cause and patch is where the operational discipline lives. A patch is what makes the equipment work for the next shift. A root cause fix is what makes the equipment work for the next two years.
For each of the three top items in the log, the root cause fix looked like this:
Walk-in refrigeration Patch cost per event: $180 Events last 30 days: 4 Total patch spend: $720 Root cause fix: $340 (new seal, proper mounting) Payback: after 2 events Ice machine Patch cost per event: $220 Events last 30 days: 3 Total patch spend: $660 Root cause fix: $1,400 (compressor replacement) Payback: after 7 events (10-14 weeks) Dishwasher Patch cost per event: $160 Events last 30 days: 4 Total patch spend: $640 Root cause fix: $520 (door gasket + heating element) Payback: after 4 events (approx 3 months)
All three root-cause fixes paid back inside 90 days. Once they were done, the same equipment did not appear in the log again for six months. The R&M line dropped 1.4 points in the following quarter and stayed there.
Preventive maintenance for the survivors
The root-cause fixes solve the recurring failures. Preventive maintenance stops the next cohort of failures before they start. This is where quarterly discipline pays outsized dividends.
Refrigeration
Coils cleaned quarterly. Gaskets inspected quarterly. Temperature probes calibrated quarterly. This is a two-hour visit per unit, roughly $150 per visit per location. On a five-location group that is $3,000 a year. The alternative is one emergency refrigeration call at Saturday dinner, which will cost you $1,200 in service and $2,500 in lost product, at a single location.
Cooking equipment
Burners cleaned quarterly, gas valves inspected annually, thermostats calibrated semi-annually. Fryer boils biweekly and oil filtration daily. The fryer discipline especially matters because oil life is directly linked to product quality, and product quality is directly linked to guest return frequency.
Ice machines and water systems
Filters changed quarterly, machines sanitized quarterly, water lines inspected annually. Ice machines are the single most common source of health department callouts and the single most expensive equipment class to run reactively.
Two hours of quarterly PM prevents twenty hours of Saturday-night emergency calls. The math is not close and it never has been.
The R&M budget conversation
The last piece is how R&M gets budgeted. Most restaurants budget R&M as a lump sum: "R&M is 1.8 percent of sales." That is not a budget. That is a projection. A real R&M budget has three lines.
One, planned preventive maintenance, quoted quarterly against a service calendar. This is the biggest line and it is the most predictable.
Two, reactive repair reserve, sized based on the trailing 12 months of unplanned events. This line should be shrinking over time as PM catches issues before they become failures.
Three, capital replacement reserve, roughly 3 to 5 percent of the total equipment book value per year, saved into a separate account so that when the walk-in compressor finally fails at year 12, you are not scrambling.
Split like that, the R&M line becomes controllable. Lumped together, it drifts.
The vendor conversation that changes everything
Once the log is running and the Pareto is visible, the next conversation is with the service vendor. Most restaurants use a rotating cast of same-day handymen and one or two contracted service vendors. That mix is wrong for cost control. The right mix is one primary service vendor per equipment class, on a quarterly PM contract with priority response, plus a small reserve for emergency-only calls.
The pitch to the vendor is straightforward. You are moving from ad-hoc calls to a quarterly PM contract across all locations. In exchange, you want a fixed per-visit rate, a two-hour emergency response window on covered equipment, and a monthly invoice summary that breaks out every visit by location and equipment. The vendor almost always says yes because ad-hoc work is more expensive to run than contract work on their side too.
The contract does three things at once. It converts an unpredictable reactive expense into a predictable planned expense, which makes the R&M budget forecastable. It creates a service-level guarantee so equipment failures do not turn into Saturday-night crises. And it produces a monthly report that tells you which equipment is actually consuming the most service hours, which is the data you need to make replace-versus-repair decisions.
Capital replacement is a plan, not a surprise
The third failure mode in R&M discipline is not budgeting for replacement. Every piece of equipment in a commercial kitchen has a useful life. Refrigeration compressors run 10 to 15 years. Cooking equipment runs 8 to 12 years. Ice machines run 7 to 10 years. Dishwashers run 6 to 10 years. Small wares wear out on their own timeline. All of this is knowable and forecastable.
The discipline is a rolling equipment inventory with age and expected replacement year for every major piece. Update it annually. Reserve capital at 3 to 5 percent of book value per year into a separate account. When the walk-in compressor fails in year 13, the money is already sitting there, the replacement is on order the same day, and the R&M line does not spike because the replacement is a capital event, not an operating expense.
Operations that skip this discipline end up doing three things badly. They defer replacements past useful life, which means more reactive repairs and higher R&M. They pay full retail on emergency replacements because they did not have time to shop. And they treat every replacement as a P&L event instead of a capital event, which distorts the R&M line and makes it look worse than it is.
Every piece of equipment in the kitchen has a known useful life. Every piece of equipment in the kitchen is going to fail on a knowable schedule. R&M creep is what happens when you do not act on either of those two facts.
What I got wrong the first time
Two lessons from doing this poorly before doing it well.
First, I used to make the log optional. General managers who cared about the log kept it. The ones who did not, did not. Result was a spotty dataset that could not surface the Pareto cleanly. The fix was making the log mandatory by tying it to petty cash. No log entry, no petty cash reimbursement. Log participation went from 40 percent to 98 percent in one week.
Second, I used to skip the nameplate photo. I thought "walk-in fridge" was enough identifier. It was not. Some locations had three walk-in refrigeration units and the log could not tell them apart. Once we started photographing the equipment nameplate on every entry, we could track which specific unit was the recurring failure. That is when the root-cause fixes actually landed on the right equipment.
The point
R&M creep is the quietest of the four cost lines because it drifts slowly and no single quarter's move is alarming. Over two years it grows into a 2-point margin drag and a service-reliability problem, and the operator running the unit cannot tell you why because there was never a log.
Fix is boring and mechanical. Log every repair for 30 days with a nameplate photo. Sort the log. Fix the three items that account for most of the spend, root cause, once. Install quarterly PM on refrigeration, cooking, and ice equipment. Split the R&M budget into planned, reactive, and capital reserve lines. Give it two quarters.
The R&M line bends back down. It stays down. The kitchen stops losing an hour a shift to broken equipment, which is a morale win nobody wrote a check for. And the next general manager who inherits the operation inherits a workflow, not a mystery.