The turnover that made you doubt everything

For eight months, your cleaner was the person you never had to think about. Turnovers happened. Guests praised the place. The word spotless showed up in reviews often enough that you stopped noticing it.

Then came the one. A ribbon of hair in the bathroom sink. A kitchen trash can that never got emptied. Towels folded in some improvised way that wasn't the way. Maybe a guest mentioned it, gently, in a message that landed in your stomach like a dropped plate.

And just like that, the story rewrites itself. Are they slipping? Have they gotten comfortable? Do they still care? You draft a text in your head — firm but fair — and you feel, underneath it, a small grief. You liked not having to worry.

Before you send it, it's worth asking a colder question: what does one bad turnover actually tell you? The honest answer is less than it feels like it does — and the reason why has a name.

Every turnover is a sample, not a verdict

A person is not a machine that outputs the same result every time. Your cleaner arrives with a different night's sleep, a different number of stops before yours, a sick kid, a car that wouldn't start, a checkout that ran forty minutes late and ate the buffer. On any given day, the quality of a turnover is drawn from a range — a distribution of possible outcomes — not stamped out from a fixed mold.

Most days land near the middle of that range. That middle is what you've been enjoying for eight months. But a distribution has tails. Every so often, by nothing more than the ordinary roll of circumstance, you get an unusually bad result. And every so often you get an unusually perfect one — the turnover where even the throw pillows look art-directed.

Here's the part that trips up almost everyone: after an extreme result, the next one is very likely to be more ordinary. Not because anything changed, but because extremes are rare by definition. This is regression to the mean, and it is one of the most reliably misread patterns in human judgment.

The flight instructors who learned exactly the wrong lesson

The psychologist Daniel Kahneman tells a story about training Israeli air force flight instructors. He'd been explaining that rewarding good performance works better than punishing mistakes, and an experienced instructor pushed back hard. In his experience, it was the opposite. When he praised a cadet for a beautiful maneuver, the next attempt was usually worse. When he screamed at a cadet for a bad one, the next attempt was usually better. So praise clearly spoiled them, and criticism sharpened them. He'd seen it a hundred times.

He had seen it a hundred times. He had just explained it wrong.

An unusually good maneuver is, statistically, likely to be followed by a more average — and therefore worse — one, no matter what the instructor says afterward. An unusually bad maneuver is likely to be followed by a better one, again regardless. The praise and the yelling were largely along for the ride. But because the improvement followed the criticism, criticism got the credit. The instructors had trained themselves, through vivid daily experience, to believe something false.

You are one frustrated text away from doing the same thing to your cleaner.

The trap sets itself

Watch how neatly it closes. You get a bad turnover — a tail event. You send a sharp message. The next turnover is back to normal, because normal is where things were always going to drift. Your brain files the sequence as cause and effect: I cracked down, and it worked.

So the next time there's a bad turnover, you crack down again. And it "works" again. Bit by bit you become a manager who believes pressure is the lever, when really you've just been taking credit for the weather. Meanwhile the person on the other end experiences something very different: steady effort, an off day that could happen to anyone, and a client who reads a single slip as a character flaw. That is precisely how you turn a reliable cleaner into a departing one — and cleaner churn is far more expensive than a bad turnover ever was.

How to tell real decline from ordinary noise

Regression to the mean doesn't say problems are never real. Cleaners do genuinely burn out, get overbooked, start cutting corners. The point is that you cannot tell which is happening from one data point. Signal and noise look identical up close. They only separate over time.

So the discipline is to stop judging turnovers one at a time and start reading them as a series:

  • A single bad turnover is noise until proven otherwise. Treat it as a fluke worth a neutral note, not a trend worth a confrontation. "Hey, looks like the sink got missed yesterday — all good, just flagging it."
  • Real decline shows up as a shift in the baseline, not a dip below it. Three or four sub-par turnovers inside a short window, where there used to be none, is a pattern. One is a Tuesday.
  • Beware your own memory. You will vividly recall the disaster and quietly forget the ninety quiet successes around it — that's the availability heuristic, the same mental shortcut that makes plane crashes feel more common than car crashes. Your gut is keeping a wildly biased scoreboard.
  • Judge the average, act on the trend. Praise the good baseline. Correct the genuine slope. Never confuse a good week with a promotion or a bad day with betrayal.

The uncomfortable truth underneath all of this is that your feeling of they're getting worse is not evidence. It's a hypothesis — and one your mind is strongly biased to accept too early.

You can't reason about a pattern you never recorded

All of this advice quietly assumes something most hosts don't actually have: a record. To know whether this turnover is worse than usual, you need to know what usual was. Not a vague impression — an actual history of turnovers you can look back across. Otherwise you're doing statistics with a memory that keeps only the worst entries and throws away the rest.

This is the real reason to build a lightweight paper trail around every turnover — a timestamp, a confirmation, a photo or two. Not to police your cleaner. To give yourself the base rate you need so that a single bad day reads as a single bad day, and a true decline can't hide inside your goodwill for months. A record turns "I feel like they're slipping" into "here are the last twelve turnovers," which is the only version of that sentence worth acting on.

Where this leaves you

Stayput exists to keep that record without turning you into a supervisor. Each turnover triggers a text to your cleaner, a quick photo confirmation back, and a restock check — and every one of those lands in a running history for the property. So when the doubt arrives, and it will, you're not searching your memory for the last time things looked off. You're looking at the actual sequence: eight good months, one hair in the sink, and the calm that comes from knowing the difference between a bad day and a bad trend.

If you'd rather judge your turnovers by the evidence than by the last thing that rattled you, you can see how it works at https://stayput.lumenlabs.works — and give a good cleaner the fair reading they've earned.