What a probation period is for, and what a wrong hire actually costs

Most employment in the United States is at will. The employer can end the relationship and so can the employee, at any point, and the only real constraint is that the reason must not be one the law prohibits.

A probation period does not alter that in either direction. It does not create a right to dismiss that did not exist, and it does not create protection after it ends.

Which raises a fair question: what is it for. The answer is that it is a management device rather than a legal one, and used properly it is a good one.

What the period actually does

Three things, none of them legal.

It sets a decision date. Without one, a marginal hire drifts. Ninety days in, nobody has decided anything, and at six months the conversation is harder because time has been read as approval. A defined review date forces a judgement while it is still cheap to act on.

It sets expectations explicitly. A written introductory period paired with what will be assessed gives a new employee something to aim at. Most people who fail early fail because nobody told them what good looked like, not because they were incapable.

It creates a documentation habit. Notes taken at thirty and sixty days are the record that supports whatever decision comes at ninety. That record matters if the decision is ever questioned, and it matters more for the far more common case where the answer is to keep the person and address one specific issue.

What it does not do

It does not remove protection against dismissal for a prohibited reason. Discrimination and retaliation protections apply from the first day, and a probation period is no defense at all.

It does not affect wage and hour obligations either. Minimum wage, overtime and required breaks apply during the period exactly as afterward, and a reduced training rate is lawful only inside limits that are specific and narrow.

Those are the rules a shorted employee would take to the Department of Labor, and several states stack their own on top of them.

And it does not create an implied contract for the period, unless you write one by accident. Language promising employment for ninety days, or listing the only reasons someone may be dismissed, can do exactly that in some states. Keep the wording to what is assessed and when, and avoid promising duration.

Structuring one that works

1. State the length and the review dates in the offer letter, along with a sentence preserving at-will status.
2. Write down three to five things that will be assessed, specific to the job. Reliability, quality of a defined output, safety practice, willingness to ask rather than guess.
3. Hold the thirty-day conversation even when everything is fine. A meeting that only happens when there is a problem becomes a signal by its existence.
4. Write two lines after each conversation. Date, what was discussed, what was agreed.
5. Decide at the end, out loud. Confirm continuation in writing. Silence is not a decision, and it is read as one.

What a wrong hire costs

The visible number is wages paid for work that did not meet the need. It is usually the smallest component.

Estimates of total cost as a multiple of salary circulate widely and vary enormously by role, so a specific figure would be an invention. What is reliable is the shape: the cost is several times the wages paid, it is concentrated in senior people's time, and it rises steeply the longer the decision is postponed.

That last point is the one that justifies the whole structure. A hire ended at week six costs a fraction of the same hire ended at month eight, and the difference is almost entirely the supervision and rework accumulated in between.

The part that saves more money than the exit

Most introductory periods end in continuation, which means the return on the process comes from the people who stay rather than the ones who do not.

Structured early check-ins catch fixable problems while they are still small: a misunderstanding about a procedure, a tool nobody explained, a shift pattern that is not workable. Each of those quietly becomes a resignation at month five if nobody asks. Asking at thirty days costs a fifteen-minute conversation.

There is a second-order effect worth naming. A workplace where new people are checked on early and told plainly how they are doing develops a reputation among the people who might work there, and in a tight local labor market that reputation is worth more than a modest difference in pay.

Small employers rarely think of an introductory process as recruitment, and in practice it functions as one.

The employers who do this well tend to report the same thing. The probation period stopped being about whether to keep people and became the mechanism by which new people got competent faster, which is a better use of it and a considerably cheaper one.

When the answer is no

Handle the ending as carefully as the hiring. Say it directly, in person, with a short and factual reason, and do not litigate it.

Pay everything owed on the schedule your state requires, which in some states means final wages on the day of separation rather than at the next ordinary pay date, and get the paperwork right, since an administrative error at the end is the most common way a routine separation turns into a claim.

Then look at the hiring rather than at the person. A hire that failed in six weeks usually failed for a reason visible in the process: a job description that described the wrong work, an interview that assessed enthusiasm rather than capability, or a reference check nobody made.

Fixing that costs nothing and is the only part of the episode with any value left in it.