Hiring Freeze: When to Use One and How to Run It

A hiring freeze is the easiest cost decision a small business ever makes, which is precisely the reason to be careful with it. Nobody loses a job. Nothing has to be announced to customers. A number moves in the right direction on a spreadsheet the same week it is decided.

The cost is real. It just arrives later and lands on somebody else. I have run a freeze and watched several more from close range, and the pattern repeats. The freeze itself is rarely the problem. The problem is everything nobody decided: which open roles are actually dead, who can approve an exception, what happens to the candidate in a final round on Thursday, and whether the manager who just lost a key person gets a backfill or gets told to absorb it.

This guide takes those decisions one at a time. It starts with soft and hard freezes, then when a freeze is the right instrument and when it postpones a layoff decision already made privately. After that come offers already extended, the announcement, the people absorbing the unfilled work, the exception process that keeps it honest, the day it lifts, and the alternatives.

I build the people and records tooling for businesses without an HR department at FirstHR. This is general information rather than legal advice, and employment law here varies by state.

What a Hiring Freeze Is

A hiring freeze is a temporary, employer-imposed pause on filling open positions and creating new ones. It changes nothing about anybody currently employed. It changes only what happens to work that has not yet been assigned to a person.

That last point, what the freeze covers, is where the confusion lives. Two people can hear the phrase and picture different things. A freeze that covers only new roles and a freeze that also blocks backfills, meaning replacements for people who leave, are separated by an enormous amount of practical pain for the people who stay.

A freeze is not a furlough, which pauses the employment of people who already work for you. It is not a reduction in force, which ends jobs. A freeze creates no final paycheck, no continuation coverage question (whether a departing employee can stay on your health plan), and no notice obligation, because nobody has experienced an employment loss.

It is also not free. It costs nothing on the payroll line and charges the difference somewhere else: overtime for the people covering the gap, contractor rates at a premium to salary, projects that slip, and eventually the cost of restarting a pipeline from zero. A freeze does not pause the clock. It moves it.

Soft Freezes and Hard Freezes

A soft freeze stops new requisitions, the formal requests to open a role, while allowing approved and in-process roles to continue through a named exception route. A hard freeze stops all external hiring, closes open requisitions, cancels scheduled interviews, and may withdraw offers that have not yet been accepted.

The distinction that matters to your team is narrower than either label suggests. It is the backfill question. A freeze that blocks new roles but funds replacements when somebody resigns is survivable. A freeze that blocks replacements too is a slow reduction in force delivered by attrition, and it should be described that way internally, because that is what people will experience.

Note the row on internal moves. A freeze holds total headcount flat, so promoting somebody into a gap and backfilling the smaller role costs less than an external hire and is frequently the fastest answer available. That makes a freeze one of the few moments where internal recruitment gets the attention it deserves.

When a Hiring Freeze Is the Right Instrument

A freeze is the right instrument when the problem is timing rather than structure. Revenue is late but contracted. A funding round is close but not closed. A large client is renewing but has not signed. In each case the current headcount is affordable at the revenue you reasonably expect, and what you need is a few months of reduced outflow.

It is also the right instrument when you genuinely do not know yet. An acquisition under discussion, a product decision that has not been made: pausing new commitments while the picture resolves is reasonable, provided you say that is what you are doing.

That quits figure is the most useful input to a freeze decision and almost nobody looks it up. Employers routinely assume attrition will shrink headcount faster than it does, then discover four months in that nobody has left and the payroll number has not moved.

Finally, a freeze is right when the alternative is a badly timed layoff. Cutting people in the month before a contract lands is expensive twice: once in severance and once in the rehiring you do at a premium. Holding the line for a quarter is often cheaper, and it is easier to reverse.

When a Freeze Is a Way of Avoiding a Decision

A freeze becomes avoidance when the current headcount is unaffordable at any revenue you actually believe in. At that point the freeze does not solve the problem. It postpones it while the cash it was meant to preserve keeps draining.

Three questions separate the honest version from the evasive one. Would the business be viable at current headcount if revenue hit the plan you would defend to a lender? Is there a specific event that resolves the uncertainty, with a date attached? And if the freeze runs a full year and nothing changes, is there a decision waiting at the end, or is the freeze the decision?

If the answers are no, no, and the freeze is the decision, what you are running is a reduction in force delivered slowly through unfilled vacancies, without the honesty or the planning that a real one gets. The people absorbing the work know. They are the ones covering two jobs while being told the company is fine.

There is a counterweight here, and it argues for the freeze. Research published in Harvard Business Review in October 2024 found that after the layoffs of 2020 to 2022, employee engagement took twelve to eighteen months to rebound. For 2023 layoffs, recovery was running longer, at eighteen to twenty-four months, and only where essential roles were backfilled (Harvard Business Review).

A layoff is not a clean reset. If a freeze genuinely avoids one, it is worth a great deal. What a freeze cannot do is avoid one indefinitely. The notice requirements and the severance question do not become cheaper for having been postponed, and a cut delivered late is delivered with less cash behind it.

Offers Already Out and Candidates Already in Process

Honor accepted offers. That is the rule I would give any small business owner, and the exceptions are narrow enough that the owner should personally sign off on each one.

Legally, most US employment is at will, so an accepted offer can generally be withdrawn before the start date without ending an employment relationship that has not begun (Cornell Legal Information Institute). That is the legal position, but not the whole picture.

Two things complicate it. Depending on the state, a candidate who resigned a job, declined other offers, or relocated in reliance on your written offer may have a detrimental reliance claim, which argues they were harmed by acting on your promise. And withdrawal patterns that fall unevenly across groups protected by anti-discrimination law create a separate exposure.

For everyone else in the pipeline, the effort you owe scales with the effort they have already given you.

The candidates you handle well during a freeze are the pipeline you restart with. Keeping them warm through a pause is cheaper than sourcing new ones later, which is the entire argument for maintaining a talent pool and treating candidate engagement as an ongoing activity rather than a campaign.

How to Announce and Run a Hiring Freeze

Say it once, in writing, to everybody at the same time, with the scope, the exception route, the expected duration, and the review date in the first paragraph. Managers hear it an hour earlier so they are not learning about it from their own teams.

The instinct in a small business is to keep it quiet and let it be noticed. That never works. People see requisitions disappear and interviews stop, and the story they construct in the absence of information is always worse than the truth. An unannounced freeze reads as a company in trouble that is hiding it. The eight steps below run from the first decision to the monthly review.

The scope decisions are the part that has to exist on paper before the announcement is written, because the announcement is just this document in prose. Fill it in, then write from it.

Externally, the footprint has to match the message. Live postings during a freeze generate applications you will never answer, and a careers page advertising eight roles that no longer exist is a slow-acting reputation problem. Take them down, or replace them with a short honest note.

Recruiters and agencies need written notice, not a phone call you both remember differently. Confirm what happens to candidates already submitted and whether any fee obligation survives the pause, because that conversation is harder after somebody places a candidate you cannot hire.

The Effect on the People Who Stay

The bill for a hiring freeze is paid in absorbed workload, and it is paid by the people you least want to lose. Every unfilled vacancy is work that still exists, and in a small business there is no bench to hand it to.

This is where the backfill refusal does its damage. A manager loses somebody to a resignation, asks for a replacement, and is told no. The team of five is now a team of four doing the work of five, with no end date, and the manager has to explain a decision they did not make and cannot defend.

Gallup research published in 2018, based on a study of nearly 7,500 full-time employees, identified an unmanageable workload as one of the five main causes of burnout. The other four were unfair treatment, lack of role clarity, poor manager communication, and unreasonable time pressure. A freeze that runs long enough manufactures all five conditions at once.

The cost of that burnout is measurable. In the same Gallup research, employees who reported burnout very often or always were 63 percent more likely to take a sick day and 2.6 times as likely to be actively looking for another job (Gallup).

There are four things that materially reduce the damage, and none of them cost money. Name what gets dropped, rather than expecting the same output from fewer people and calling the gap resilience. Give the absorption an end date tied to the freeze review, so it is a stretch rather than a new permanent job description.

The other two cost only attention. Recognize the extra work explicitly, in the same channel where the freeze was announced. And check on the people carrying the most, because burnout is quiet until it is a resignation.

The thing that reliably makes it worse is refusing to acknowledge it. A team quietly absorbing a departed colleague’s workload while leadership talks about efficiency gains will draw its own conclusions about how the company sees them, and those conclusions show up in retention a quarter later.

Exceptions and the Approval Process

A freeze without a written exception process becomes a freeze that leaks. Requests get approved by whoever is asked at a good moment, the rules turn out to be different for different teams, and within two months the freeze exists only in the announcement email.

The point of the process is not to make exceptions hard. It is to make them visible and consistent, so that a manager refused a backfill can see that the manager who got one met a published test rather than caught the owner in a better mood.

Attach a real cost figure to every request. Base salary understates the commitment once employer payroll taxes, benefits, equipment, software seats, and sourcing spend are counted. Only the full figure tells you whether an exception eats the savings the freeze was meant to produce.

Two categories deserve automatic exceptions and should be named in the announcement. The first is roles required by law, licensing, or a customer contract, where leaving the position empty creates an exposure larger than the salary. The second is single points of failure, where one person leaving takes a system, a certification, or a client relationship with them and there is no cover.

Keep the requests, the refusals, and what happened to the work in the same file. The refusals are the half that never gets recorded, and they are the half that tells you what the freeze actually cost.

How to Decide When to Lift It

Lift the freeze against the condition that caused it, not against the calendar and not against how things feel. If it was called because cash coverage fell below a threshold, the trigger for lifting is that threshold being restored and held for a defined period.

Write the trigger down at the moment you announce the freeze. Deciding it afterward means the decision gets made by whoever is most persistent, and the freeze ends with an argument instead of a plan. A written trigger also gives you something honest to say each month when people ask, even when the answer is that nothing has changed.

Lift in stages rather than all at once. Reopen the roles the exception process flagged most often, then the roles covering the heaviest absorbed workload, then genuine growth roles. Reopening everything simultaneously produces a hiring surge nobody has capacity to run, which is how a company goes from a freeze to a set of bad hires in one quarter.

Build in the lag. A pipeline shut down for two quarters does not restart on the day of the announcement, and approving a role is the start of a hiring cycle rather than the end of a problem. Look at your own time to hire and then plan the reopening backward from when you need the person productive. That is the same arithmetic that belongs in headcount planning in the first place.

Tell people it has lifted with the same visibility you used to announce it. A freeze that is quietly abandoned teaches the team that announcements from leadership expire without notice, and the next one gets taken less seriously.

The Alternatives Worth Considering First

A freeze is one of several instruments for holding labor cost flat, and it is not always the best one. The three most useful alternatives for a small business are managed attrition, reduced hours, and converting work to contractors.

Managed attrition is the quiet default and deserves more deliberation than it gets. Deciding in advance which roles get backfilled turns random departures into a plan, and it is kinder than deciding role by role while a manager waits.

Reduced hours works when the work has genuinely shrunk. For hourly staff it is straightforward, subject to state notice requirements and any predictive scheduling rules that require advance notice of schedule changes. Cutting hours also drops some people below the threshold where they qualify for benefits, so check your plan against part-time hours first.

Salaried exempt employees are where reduced hours gets dangerous. Reducing a schedule and docking pay by the day puts the overtime exemption at risk, the same trap that makes partial-week furloughs so expensive. Where an employee is ready, willing, and able to work, no deduction may be made for time when work is not available (DOL Fact Sheet 17G).

Contractor conversion is the alternative most often used badly during a freeze. It is legitimate for project work, specialist skills, and genuinely temporary demand, and it is the fastest way to cover a gap that has a defined end.

Used as a workaround, contractor conversion only holds the headcount number down while the same person does the same job under the same supervision. That is a classification problem with tax and wage-hour consequences, and the Department of Labor decides it on the economic reality of the relationship rather than on the contract (DOL Fact Sheet 13).

Where a Salary Freeze Fits

A salary freeze suspends raises, merit increases, and promotion bumps while everyone keeps their current pay. It is the sibling instrument to a hiring freeze: one holds the cost of people you have not hired yet, the other holds the cost of the people already on the payroll. Businesses run both together more often than either alone.

The two freezes spread their cost in opposite ways. A hiring freeze is invisible to most of the team and lands hard on the few absorbing unfilled work. A pay freeze lands lightly on each person and is visible to all of them the day it is announced, which makes it easier to explain and much harder to contain once your strongest people start comparing offers.

Check two mechanical points before you announce one. First, a frozen hourly rate still has to clear the minimum wage where the work is performed, so a state or local increase overrides the freeze for anyone sitting near the floor (DOL state minimum wage table).

The second point is the end condition. A pay freeze needs one just as a hiring freeze does, because a pay freeze nobody remembers to lift is a pay cut delivered slowly by inflation.

Whichever instrument you choose belongs in a plan rather than a reaction. That is the argument for doing workforce planning before you need it, so a cash squeeze produces a decision you had already thought about rather than an email sent at nine at night.

Where Small Employers Get This Wrong

Six patterns come up again and again, and the first two account for most of the damage I have seen.

The first is not defining backfills. A freeze that has not said whether replacements are funded will be interpreted differently by every manager, and the argument arrives the first time somebody resigns. It is a one-sentence decision that saves a month of friction.

Second comes leaving candidates to work it out. Silence after a final round is the version of this that costs most, because those candidates are the exact people you will want to call when the freeze lifts, and they will remember.

Announcing a hard freeze and running a soft one is third. The first approved exception under a freeze that promised no exceptions destroys the credibility of the whole thing, and every subsequent refusal reads as favoritism.

The fourth is setting no end condition. A freeze with no trigger and no review date persists until somebody senior breaks it out of frustration, which is the worst available ending and teaches everyone that rules erode with pressure.

Pretending the work vanished is fifth. Headcount going flat while output targets stay identical is a decision to spend people, and the people know it even when nobody says it out loud.

The last is using a freeze to postpone a layoff you have already decided on. It burns the cash that would have funded a decent severance, extends the uncertainty for everybody, and converts a hard piece of news into a broken promise. If the answer is a reduction in force, doing it properly and early is better for the people leaving and for the people staying.

Frequently Asked Questions