Five hours is not five hours
The window in the form runs from an 11:00 am checkout to a 4:00 pm check-in, which is 300 minutes. Fifteen of those go on loading the van and getting to the first door, so what is left to clean in is 285. A property takes 1 h 55 min on site once the 20 minutes of restocking, checking and photographing is counted with the 1 h 35 min of cleaning, and moving to the next one costs another 18. Two properties therefore take 4 h 23 min of the window and fit. Three take 6 h 36 min and do not, and no amount of hurrying closes a gap that size.
That is the whole constraint, and it is why the answer to a late day is usually a different number of teams rather than faster cleaners. Set the drive between properties to zero, as though every unit were in one building, and a team still only fits two: the finish moves from 5:36 pm to 5:00 pm and the team count does not move at all. The cleaning time is doing almost all of the work.
The team with the extra property sets the day
Five turnovers across two teams is three and two, not two and a half. The team with three finishes at 5:36 pm and the team with two finishes at 3:23 pm with 37 minutes to spare, and the day is judged by the first of those. This is the quiet arithmetic behind a lot of bad afternoons: the average looks comfortable, the worst case is an hour and a half past check-in, and nobody notices until a guest is at the door.
The table lower down the page walks the team count from one upward. One team finishes at 10:02 pm with three properties late. Three teams finish at 3:23 pm with none. Between those two the number of late properties falls from three to one to zero, and the third team is the one that buys the whole day.
What the two times on your listing are worth
With the teams you have, the latest checkout that still clears the day is 9:24 am and the earliest check-in that works is 5:36 pm. Both are stated as clock times rather than as a slack figure because that is the form the decision takes: you either change what the listing says or you change who is working. Moving checkout an hour earlier is not enough here, which is worth knowing before you try it.
The last row is the honest one. A single clean running 25 minutes over pushes the finish to 6:01 pm, wherever in the run it happens, because everything behind it moves with it. A day planned with less slack than one bad clean is a day that depends on nothing going wrong.
Questions people ask
How many cleaners do I need for five same-day turnovers?
For the day in the form — 11:00 am checkout, 4:00 pm check-in, 1 h 35 min of cleaning plus 20 minutes of restocking per property, 18 minutes between them and 15 from base — one team fits two properties, so five turnovers take three teams. With two teams the last property is finished at 5:36 pm, an hour and 36 minutes after check-in.
Would cutting the drive time between properties fix a late day?
Rarely, and the page will show you. Setting the drive to zero on the default day, as though every unit were in the same building, moves the finish from 5:36 pm to 5:00 pm and still leaves one property late and three teams needed. Two hours of cleaning per property against a five-hour window is the binding constraint, and eighteen minutes of driving is not.
What is the latest checkout time I can advertise?
9:24 am for the day in the form, with two teams and three properties on the busiest one. That is the time at which the last property is finished exactly at check-in, with no margin. The page prints it alongside the earliest check-in that works with the same crew, because those two are the levers that cost nothing except what they do to bookings.
Does this account for a clean that goes wrong?
Only through the one row that asks it to. A property running 25 minutes over pushes the finish to 6:01 pm, and it does that whether the bad clean is the first of the run or the last, because everything after it waits. The page does not model how often that happens — that is your own record — but it does show that a day with less than one bad clean of slack has none.