How to Assign Cleaning Crews to Multiple Locations: 9 Tests

A route that looks clean in a spreadsheet falls apart at 11:40 pm on a Tuesday, when the crew is still burnishing Building C and Building D's alarm auto-arms at midnight. You rarely lose a multi-site account over a dusty ledge. You lose it because the fourth stop was never reachable in the hours you sold.

Assigning crews across several buildings is not a scheduling chore. It is an evaluation problem: you are testing whether a proposed crew-to-site pairing survives contact with locked doors, drive time, call-outs and payroll law.

Assign crews to multiple locations by testing the plan before you publish it: confirm every site's access window, add up task minutes from your own timed rates, add travel and lock-up time, then keep 10% to 15% of the shift as slack. If one site on the route fails, the whole route fails.

Below are nine tests, ordered so the cheap ones eliminate bad plans before you spend a night in the truck proving it.

The nine tests, in the order that kills a bad route fastest

Run tests 1 through 3 at your desk with a calendar and a calculator. Test 4 needs a map. Tests 5 through 7 need your roster and your key log. Only tests 8 and 9 require you to actually run the route.

Test 1: Do the access windows actually overlap?

Pull the service window out of every contract on the proposed route and write them on one line. Building A, 6:00 pm to 5:00 am. Building B, after the last patient leaves, typically 7:30 pm. Building C, tenant-occupied until 9:00 pm on the third floor.

Pass: there is a continuous block of time in which one crew can visit every site in sequence without waiting outside a locked door.

Fail: two sites demand the same narrow window, or a site opens after the crew's shift would need to end. This is fatal and it is free to detect. Do not proceed to task-time math until it clears.

Test 2: Does the route fit the shift? The 10% to 15% slack rule

Total the task minutes for each site using your own timed production rates, not a vendor's brochure. If you have never timed your own crews, ISSA publishes task-level cleaning times you can use as a starting point until you collect real data.

Then add the three things operators consistently forget: travel between stops, site opening and lock-up (unlock, disarm, cart out, trash to dock, rearm), and a slack buffer.

Pass: planned work consumes 85% to 90% of the available window, leaving 10% to 15% as slack. That buffer absorbs a spilled toner cartridge, a conference room left in use, and traffic.

Fail: under 10% slack the crew starts cutting detail work in week three, usually restrooms and high dusting. Over about 20% slack on an hourly crew, you are paying for idle time and should tighten the route or add scope.

Worked example: three cleaners, four buildings, the Northgate route

Assumptions, all illustrative: your timed rate on light-density office is 3,000 sq ft per cleaner-hour, and 2,200 sq ft per cleaner-hour in the clinic because of exam room protocols. Crew of three, paid 6:00 pm to 1:00 am, a 7-hour window.

• Building A: 22,000 sq ft office, 5 nights. 22,000 / 3,000 = 7.33 cleaner-hours per visit.
• Building B: 9,000 sq ft dental clinic, 3 nights. 9,000 / 2,200 = 4.09 cleaner-hours.
• Building C: 14,000 sq ft office suite, 5 nights. 14,000 / 3,000 = 4.67 cleaner-hours.
• Building D: 6,500 sq ft office, 2 nights. 6,500 / 3,000 = 2.17 cleaner-hours.

Tuesday hits all four sites: 7.33 + 4.09 + 4.67 + 2.17 = 18.26 cleaner-hours. Split across three cleaners, that is 6.09 hours each of pure task time.

Now the overhead. Three hops at 12 minutes each is 36 minutes, and four site openings and lock-ups at 10 minutes each is 40 minutes. That is 76 minutes, or 1.27 hours, per person. Total per cleaner: 7.36 hours against a 7-hour window.

Tuesday fails before anyone turns a key. Move Building D to Monday and Wednesday, nights when the clinic is not serviced, and Tuesday becomes 16.09 cleaner-hours / 3 = 5.36 hours, plus 24 minutes of travel and 30 minutes of open and close, for 6.26 hours. Slack is (7 - 6.26) / 7, or 10.6%. That passes.

Test 3: Do you have to pay for drive time between buildings?

Under federal rules, travel from one job site to another during the workday is hours worked and must be paid. The ordinary home-to-first-site commute generally is not. Several states, California among them, apply stricter definitions of controlled travel, so check your state before you assume.

Pass: every minute between the first punch and the last punch is either budgeted as paid work or eliminated from the route.

Fail: you priced the account on cleaning hours only and expect crews to drive between sites off the clock. That is unbudgeted payroll at best and a wage claim at worst. It also silently pushes overtime into weeks you thought were clean.

Test 4: How far apart can the buildings be?

The useful ratio is drive minutes divided by on-site minutes for each stop. A 20-minute drive to a 25-minute strip mall office is an 80% overhead ratio. You are running a courier service that occasionally mops.

Pass: no single hop costs more than roughly a quarter of the on-site time at the stop it leads to, and the whole route sits inside a radius your crew can cross in the slack you budgeted in Test 2.

Fail: a stop where drive time approaches or exceeds clean time. Fix it by re-sequencing so the outlier is first or last, by moving that building to a different route, or by converting it to a day-porter visit tied to a nearby anchor account.

Test 5: Is every cleaner cleared for every site on the route?

A multi-site route quietly assumes interchangeability. Then the crew hits a dialysis suite, a food production floor, or a tenant that requires a background check on file with building security.

OSHA requires bloodborne pathogens training at the time of initial assignment for employees with reasonably anticipated occupational exposure, and hazard communication training for the chemicals they handle. That training has to exist before the first shift, not after the first incident.

Pass: for each site on the route, every assigned cleaner has the required training, badge, background clearance and equipment certification on file, with expiration dates you can see.

Fail: a certification gap anywhere on the roster means that site has fewer eligible cleaners than you think, which will detonate in Test 6.

Test 6: Who covers the route when someone calls out?

Single-site accounts fail gracefully when a cleaner calls out. Multi-site routes fail catastrophically, because the missing person was carrying an eighth of four buildings, and the remaining crew has no slack to absorb it.

Pass: every building on the route has at least two cleaners on the roster who have worked it inside the last 60 days, hold the required clearances, and can get in the door. Two-deep, verified, not theoretical.

Fail: a site where exactly one person knows the alarm sequence, the trash route or the tenant quirks. That is not a schedule, it is a hostage situation. Fix it with a documented site orientation and a planned rotation of the backup cleaner through that building once a month.

Test 7: Who holds the keys, codes and badges?

Access is the constraint people evaluate last and regret first. Map every site on the route to a physical access method and a named holder.

• Keys and fobs: who signs them out, where they live between shifts, what happens at 9:00 pm when the holder is stuck in traffic.
• Alarm codes: individual codes per employee if the panel supports it, so a bad disarm points to a person and not to "the cleaners".
• Building security badges: lead times can run days or weeks. A badge request submitted the week you start is a route that misses its first two nights.
• Termination path: the exact steps to revoke access at every site within 24 hours of a separation.

Pass: every site has a primary and a backup access holder, and you can revoke access everywhere in one day.

Fail: shared codes, unlogged keys, or a single ring that opens four clients' buildings riding in one person's glovebox.

Test 8: Can you prove the crew was at each stop?

On a multi-site route, "did anyone service Building D on Thursday?" is the question that decides a renewal. Punch data tied to a location answers it in seconds. Memory does not.

Geofenced clock-in is the standard tool, and the radius matters more than operators expect. Set it too tight and phone GPS drift generates false rejections on a rainy night in a parking garage. Set it too wide and the punch proves nothing. Note that most geofence systems capture a location point at clock-in and clock-out, not a continuous trail.

Pass: one week of punch data shows arrival and departure at every stop, in sequence, with fewer than a handful of manual corrections.

Fail: repeated "outside the zone" errors, or supervisors editing punches nightly. Widen the radius for that specific site rather than training your crew to ignore the system.

Test 9: Does the route still make margin after drive time?

Run this last, because it needs the outputs of Tests 2 and 3. Take the Northgate route and price it.

Weekly task hours: Building A 7.33 x 5 = 36.65, Building B 4.09 x 3 = 12.27, Building C 4.67 x 5 = 23.35, Building D 2.17 x 2 = 4.34. Total 76.61 cleaner-hours.

Now add the overhead a naive estimate ignores. Roughly 2.5 crew-hours per week of site opening and lock-up plus about 2 crew-hours of travel, times three cleaners riding the clock, is about 13.5 cleaner-hours. Weekly total: 90.1 cleaner-hours.

Assume, illustratively, a fully loaded cost of $22 per cleaner-hour, covering wage, payroll taxes, workers compensation and paid time off, and monthly revenue of $17,000 for the four buildings. Monthly hours are 90.1 x 4.33 = 390, so direct labor is $8,584, or 50.5% of revenue.

Price it on task hours only and you would have used 331.7 hours and $7,297, which reads as 42.9%. The 7.6 point difference is windshield and door time. That gap is where multi-site routes quietly lose the money the operator thinks they are making.

Pass: direct labor including travel and lock-up lands at or below your target, and many janitorial operators plan around direct labor consuming no more than half of recurring revenue.

Fail: the route only works on paper. Your options are re-sequencing, shifting a site to another route, renegotiating frequency, or repricing at renewal. Discovering it now costs a spreadsheet. Discovering it in month seven costs a year of payroll.

Frequently asked questions

Do I have to pay cleaners for drive time between buildings?

Yes, under federal rules. Travel from one job site to another during the workday counts as hours worked and must be paid, per 29 CFR 785.38. The regular commute from home to the first site and from the last site home generally does not. Some states apply broader definitions when you control the travel, so verify your state's rule before you build the route.

How many buildings can one crew realistically cover in a night?

There is no fixed number, because the limit is minutes, not stops. Run the math instead: total task hours divided by crew size, plus travel and open and lock-up time per person, must land inside 85% to 90% of the shift window. In practice, tightly clustered small offices support four or five stops. Sites more than 15 minutes apart usually cap out at two or three.

Should I keep crews fixed to the same buildings or rotate them?

Keep the primary assignment fixed and rotate the backup. Fixed crews learn tenant quirks, alarm sequences and where the freight elevator key hides, and clients value seeing the same faces. But a route with zero rotation creates single points of failure. Cycling the designated backup through each building once a month buys resilience without losing continuity.

What geofence radius should I use for a strip mall with a shared parking lot?

Start around 150 to 200 meters so the whole lot is covered, then accept the tradeoff: that radius will also cover neighboring tenants. If proof of presence at a specific suite matters, pair the punch with a photo-based inspection or a checklist item completed inside the space. A geofence proves proximity, not that a restroom got cleaned.

How do I cover a call-out on a five-stop route without blowing the schedule?

Decide the triage order before it happens. Rank the route's sites by contractual risk and detail sensitivity, then pre-authorize your supervisor to reduce a specific low-risk building to trash, restrooms and entry spot-mopping on short-staffed nights. Notify that client the same night. A planned, disclosed reduction survives an account review. A silently skipped building does not.

Where CleanTrack360 fits

The nine tests above run on paper. Keeping them true week after week across a dozen buildings is where software earns its keep. CleanTrack360 handles the moving parts: drag-and-drop scheduling with recurring shifts and work orders, geofenced GPS clock-in and clock-out that runs in the phone browser with a default 150 m radius you can configure per location, quality inspections with photo evidence and automatic scoring, and reports you can export to CSV to rerun your labor-to-revenue math. Locations can be loaded in bulk by CSV, and location-based message channels keep each building's crew on one thread. There is no published native app yet, so crews work from their phone browser today.

Plans are Starter at $99 per month for up to 5 team members, Pro at $199 for up to 20, and Business at $249 for up to 50, priced per plan rather than per user. There is a 14-day free trial and no credit card required, which is enough time to run one route through Test 8 and see what your punch data actually shows.