Night Shift Wages vs. a Fixed Robot Payment: A Worked Example

Which costs belong on each side of the comparison?

A night crew's real cost is not its hourly rate. It is every paid hour on the night shift, any premium the employer adds for working nights, the payroll burden on top of wages, and the expense of replacing people who quit. A fixed robot payment has to be set against that whole bundle, measured over the same 36 months.

On the robot side, the main line is the fixed monthly payment under a three year lease. The question is what else the site still pays for: supervision, exception handling, and the staff who respond when something goes wrong. Those costs do not disappear because a machine is on the floor, so a fair worked example puts them on the page too.

What inputs must a manager gather first?

Start with hours, coverage and turnover. Each input changes the answer, and each should come from your own records or a named source.

• Paid night hours per shift, shifts per week and weeks per year, which together set total crew hours.
• Coverage area and task list. A floor scrubber covers aisles and corridors, while a crew may also clean restrooms, empty bins or reset stations.
• Headcount needed to cover those hours, including breaks, sick days and vacancies.
• Base wage and night differential. A flat dollar premium and a percentage premium scale differently as wages rise.
• Turnover rate and replacement cost per departure, each tied to a named source.
• Exception hours, meaning the time someone spends clearing a jam, a spill or a blocked route.

How big is the wage line on a night crew?

The Bureau of Labor Statistics put the median hourly wage for janitors and building cleaners at $17.71 in May 2025, or $36,840 a year. The bottom 10% earned under $13.77 an hour and the top 10% earned over $24.17. The median was $17.21 for workers in services to buildings and dwellings and $18.35 in healthcare and social assistance.

A vendor analysis of BLS data from Nassco reports that the median rose 26.7% from $13.98 in May 2020. That is a vendor calculation. Nassco's own notes say BLS figures are national benchmarks, not facility costs, so local wage data and your actual payroll should replace them in the model.

Night differentials add to that base. Paychex says they most commonly run 5% to 15% of base pay, or about $1 to $3 an hour. Nightowling's 2026 data puts hospitality at 5% to 10% or $1 to $3 an hour, and warehousing and logistics at 5% to 10% or $1 to $2.50 an hour. Shift Calculator's undated guide gives 10% to 15% for retail and office and 15% to 30% for manufacturing and logistics. Nightowling also cites the MRA 2025 Industrial and Production Trades Survey, which found 84% of industrial and production employers pay differentials as a flat dollar amount. Check which form your own pay policy uses.

What does turnover add to the crew side?

Turnover is where the sources disagree most, and that disagreement is useful to a manager. Opora Supply's BLS-based structural figure is roughly 14%, derived from 351,300 openings against about 2.45 million employed. Opora also calls the 75% to 200% range a secondary-source estimate with no federal publication behind it.

Industry estimates for frontline turnover run much higher. ISSA's October 2026 article says estimates put janitorial frontline turnover near 200%, at roughly $3,500 per departure, and it labels that an industry estimate rather than a federal figure. StealthAgents gives 150% to 200% at operators with many sites and $4,800 to $7,200 per replacement, which it attributes to SHRM or an itemized estimate. Opora puts replacement at $5,000 to $11,000 as an industry consensus estimate.

Use one turnover rate and one replacement figure, state where each came from, and then run the crew line at the low and high ends. Opora also gives a rule of thumb: a $1 an hour raise costs about $2,600 a year per full time worker after burden. That single number shows how quickly a wage step compounds across a three year term.

How does a fixed three year payment fit into the math?

A fixed payment is simple to model because it does not move with wages, turnover or a bad month of sick calls. Over a three year term it is 36 monthly figures that sit beside the crew's annual wage bill in the same spreadsheet. The payment itself is quoted only after a site assessment, so the example should use that quote, not a number from a brochure.

Service Robot Co. leases robots only, through financing on a three year lease with one fixed payment quoted after a site assessment. We are OEM neutral. We select the right machines across manufacturers, then deploy, integrate, train and service every unit through our nationwide US engineer network, so the lifecycle has one vendor and one number to call.

What does the comparison leave out?

The biggest gap is supervision. A robot still needs someone who reviews run logs, clears a stuck unit, confirms the route was completed and decides what to do when a spill sits in an aisle. That time may come from a lead, a manager or a remote operator, and it belongs on the robot side of the sheet.

Exception handling is the second gap. A crew member steps around a pallet, bags a broken bottle or moves a rope barrier without thinking. A robot needs a defined response for each exception, and the cost of a slow response is hard to see until the floor is wet at 5 a.m. Write response times into the plan and price the hours that cover them.

Coverage is the third gap. A crew can absorb a sick call by shifting people around, while a robot needs a backup unit or a repair window. Ask how backup coverage and emergency response work before you treat the fixed payment as the full cost of keeping floors clean overnight.

How should the worked example be built?

Build two rows on one page and keep the same 36 month window for both. Each row should show its assumptions, so a reviewer can see which input drives the gap.

• Crew row: total paid night hours, base wage, differential, payroll burden, and a turnover line using one sourced rate.
• Robot row: the quoted fixed payment over 36 months, plus the supervision and exception hours you expect each week.
• Non floor tasks: restroom resets, trash or station work that the robot will not do. Keep those hours on the crew row or reassign them.
• Sensitivity: rerun the crew row at the low and high turnover figures and with one wage step, so the spread is visible.
• Result: compare totals and name the one or two assumptions that would change the decision.

How much weight can the sourced figures bear?

Treat each sourced figure as a range, not a verdict. Government wage data is the strongest input here, and vendor analyses such as the Nassco and StealthAgents figures are useful but come from firms that sell into this market.

Where the example depends on one contested number, say so in the write up. A transparent model with a visible range is more persuasive to a finance reviewer than a single confident total.