A probation period does not make anyone easier to fire. In nearly every state, employment is already at will, and calling the first ninety days "introductory" changes almost nothing about your legal position. What the period actually buys is a deadline: a date on the calendar by which somebody has to write down whether this is working. Employers who have run this cycle dozens of times know the real cost is not the policy language. It is what happens when nobody honors the date.
If you are reading this because you let one slide, you are in ordinary company. The recovery is mechanical. But the next hire is worth structuring differently, and there are three common structures with genuinely different price tags.
Option one: an in-house introductory period
You hire the person directly, set a review date at 30, 60 or 90 days, and hold a benefits waiting period until the end of it. This is the cheapest option on paper and the most expensive one when it fails.
The visible cost is wages plus payroll taxes for the trial window, plus the supervisor hours spent training and checking. The invisible cost is everything you sink into someone before you admit the answer is no. Equipment issued. Credentials created. Client introductions made and then unmade. A team that reorganized around a person who is about to leave.
Then there is unemployment insurance. State unemployment tax is experience rated, which means separations charged to your account can raise the rate you pay on every employee for years afterward. One short, poorly documented separation is not a catastrophe. A pattern of them is a line item you will notice. The Department of Labor oversees the federal-state unemployment insurance framework, and the state agency that administers your account is the one that decides how a separation is coded.
What makes this option work is discipline about the date. The employers who get value from it write the review into the calendar the day the offer is accepted, and they write one page at each checkpoint: what the person was asked to do, what they did, what changes next. That page is the whole product.
Option two: temp-to-hire through a staffing agency
The agency employs the worker, handles payroll and unemployment exposure, and you pay a bill rate that runs meaningfully above the hourly wage. Many agreements add a conversion fee if you bring the person on directly, sometimes waived after a set number of hours worked.
You are paying a premium to move the decision off your books. For roles where you genuinely cannot predict fit until you have watched the work (warehouse, call center, entry-level administrative, seasonal surge), the premium is often worth it, because the alternative is a stream of short separations charged to your account. For a senior or specialized role, the math turns quickly. The bill rate premium over several months plus a conversion fee can exceed what a recruiter would charge to fill the seat once.
Read the conversion clause before the first day, not at week ten. The two things that drive the cost are the length of the buyout period and whether hours worked count toward waiving the fee.
Option three: a paid trial project on a defined scope
A bounded, paid piece of work with a start, an end, and a deliverable. Useful for design, writing, bookkeeping, analysis, and trades where the work product speaks for itself.
The cost is small and highly visible: you pay for the project, you keep the output, and you learn how the person communicates when something goes sideways. The risk is classification. If you direct the hours, supply the tools, and control the method, you may have an employee regardless of what the invoice says, and wage-and-hour rules under the Fair Labor Standards Act apply. Keep trials genuinely project-shaped, or pay the person as a short-term employee and skip the argument.
This option cannot test what a long relationship tests. It tells you nothing about how someone handles month four.
What actually drives the cost in all three
Three things, in order.
- Time to decision. Every week past the point where you knew is pure cost. Experienced managers usually know by week three and spend weeks four through twelve hoping to be wrong.
- Documentation quality. A separation with three dated notes is a short conversation. The same separation with nothing written becomes a contested unemployment claim, an internal argument, and occasionally a lawyer's hour.
- Whether the role was defined before the posting went up. Most failed probations are failed job definitions. If nobody can state the three outcomes the person owns, no review date will save it.
Recovering from one you let run long
Set the date now, retroactively. Write the one page you skipped, covering what was asked and what happened, and give it to the person with a specific window and specific measures. Thirty days is usually enough. You have lost the clean early exit, but you have gained real information, and a person who improves under a written plan is often a better long-term bet than one who never needed one.
Then put the review date in the offer template so the next one runs itself.