Corporate — Field report TQL-BUS-981
Two Versions of the Same Ninety Days at a Plumbing Company's Dispatch Desk
Same role, same hire, same ninety days. What separates a probation that answers a question from one that just runs out the clock is gathered before day one.
A probation period either answers a question or it doesn't. Most of them don't, and the reason is almost always that nobody wrote the question down before the hire started. Ninety days pass, the new person is still there, nothing catastrophic happened, and the probation ends by expiration rather than by decision. That is a barely adequate probation. It cost you three months of payroll and told you roughly what a coin flip would have.
To make this concrete, take one narrow case: a residential plumbing company with six trucks hiring its first dedicated dispatcher. Until now the owner's wife has been answering the phone between invoices. The role is scheduling, intake, routing, parts calls, and the customer conversation when a tech is running late. It is not a complicated job to describe and it is a very difficult job to evaluate by feel, because the person doing it badly and the person doing it well both look busy.
What the two probations have in common
Both versions run ninety days. Both are at-will, as nearly all private employment in the United States is outside of Montana, meaning the probation label does not create a contract or restrict either side. Both pay the same hourly rate. Both end with a conversation. The Department of Labor oversees the wage, hour, and recordkeeping obligations that apply to the hire from the first day worked, and neither version changes any of that. A probationary employee is an employee.
So the difference is not legal. It is informational. One version is designed to produce evidence, and the other produces an impression.
The gathering that happens before day one
Here is where the two paths separate, and it happens before the new dispatcher has walked in. The adequate version requires the owner to decide, in advance, what a good dispatcher at this company actually does differently from a poor one. Not what a dispatcher does. What the good one does differently.
For this shop, the owner sat down with his two senior techs and came out with four things:
- A job gets booked with enough information that the truck arrives with the right part more often than not. The techs could name the failure: "water heater problem" written on the ticket instead of brand, age, gas or electric, and whether it's leaking.
- The day's route doesn't zigzag across the county. The techs could point at specific mornings where two back-to-back calls were thirty minutes apart in opposite directions.
- When a tech is going to be late, the customer hears it from the office before the customer calls in.
- Emergency calls get triaged. Not every "it's flooding" is flooding.
Those four items are not a performance review template. They came out of complaints the business already had. That is the part worth copying: the measures were pulled from existing friction, not invented from a job description. The owner then spent an hour pulling baseline numbers he already had sitting in the scheduling software and on the invoices: how often a tech made a second trip for parts, and what the previous three months of callbacks looked like. Rough numbers, taken honestly, from before the hire.
The barely adequate version skips all of that. It says: let's see how she does.
The same ninety days, side by side
| Barely adequate | Good | |
|---|---|---|
| Before day one | Offer letter, start date, desk cleared | Four named measures from real complaints, plus baseline numbers from the prior quarter |
| Week one | Shadowing, "ask if you need anything" | Same shadowing, plus the four measures explained out loud and put in writing |
| Day 30 | Nothing scheduled | Thirty minutes. One measure is already fine, two are not, one can't be judged yet. Says which is which |
| Day 60 | Nothing scheduled | Thirty minutes. Checks whether the day-30 items moved, and what the office needs to supply |
| Day 90 | "You're doing great, welcome aboard," or an awkward surprise | A decision with three months of specifics behind it, in either direction |
| What it produced | An impression | Evidence, and a trained dispatcher either way |
Notice what the day-30 conversation does in the good column. It separates three categories that the barely adequate version mashes together: things that are working, things that are not, and things you genuinely cannot assess yet. That third category matters more than people expect. Emergency triage, in a six-truck shop, might come up four times in a month. One bad call out of four is not a pattern. Writing "not enough data yet" is an honest answer and it protects a good hire from a bad judgment made on thin evidence.
The conversation at day 30 is the whole product
If a probation period has one job, it is to make the day-30 conversation possible and useful. Everything gathered beforehand exists to let a manager say something more precise than "you seem to be settling in."
The precise version sounds like this: on the parts question, you're already better than what we had, the techs have noticed, keep doing it. On routing, three mornings in the last two weeks sent Marco across town and back, and I think that's because nobody showed you the zone map, which is on me and we'll fix it Thursday. On the late calls, this hasn't happened yet and I want you making that call before the customer does. On emergencies, we've had four and I'd like more before I say anything.
That takes four minutes to say. It is a different thing entirely from a performance review, because nothing is being graded. It is a status report on four specific questions, delivered early enough that the answers can still change. A dispatcher who hears it on day 30 has sixty days to move. A dispatcher who hears it on day 90 has a surprise and a grievance.
What the good version buys when the answer is no
Both probations can end in a parting. The difference is what the parting costs. In the barely adequate version, the owner has a vague sense that it isn't working, no written record, and a conversation that will feel arbitrary to everyone involved, including the techs who will hear about it. In the good version, there are two dated conversations on file, documented at the time, each naming specifics and each identifying what the business owed the employee in support. If the separation is challenged, or a state unemployment claim raises questions, that record is the thing that exists.
And when the answer is yes, which it usually is, the good version has produced something extra: ninety days of concrete feedback means the dispatcher already knows what the job is for. The business got a trained employee instead of a surviving one.
The preparation here was about two hours of work spread over a week, done by a man who did not have two spare hours. He had the complaints already. He had the invoices already. He just hadn't written down what he was going to be watching for, and that, far more than the length of the probation or the wording of the offer letter, is what decides whether the ninety days were worth anything.