Boomerang Employees: The Hidden Cost of Cutting Too Deep

Last updated: August 5, 2026

By Robert Ardell, Co-Founder and Strategic Advisor, KORE1

Boomerang employees are workers who leave a company and later return, and by March 2025 they made up 35 percent of all new hires in ADP’s payroll data, the highest share since 2018. Most articles on the subject treat that as a feel-good story about loyalty. It is not. A large and growing share of those returns are companies buying back people they cut nine months earlier, at a premium, after discovering the work did not leave with them.

The cleanest public example is one you already saw.

In February 2025 the FDA terminated roughly 700 probationary employees. More than 220 of those came out of the medical device center, about a tenth of that program. Within about a week, per Associated Press reporting, entire teams of device reviewers got calls and emails telling them their terminations had been rescinded effective immediately. Not restructured. Rescinded. Whole teams of five or more, back at their desks, in the same seats, doing the same reviews.

Set the agency aside. I have no interest in the politics and this is not a piece about government. What interests me is the operating pattern, because I have watched private companies run the identical play maybe forty times since we started KORE1 in 2005, and the mechanics never change. Somebody models a reduction from a spreadsheet. The spreadsheet knows headcount and salary. It does not know that one of those names is the only person who understands the billing integration, and nobody finds out until the following quarter close.

Then the calls start. By then the person has an offer somewhere else and a higher number. It always goes up.

I should say plainly where I sit. KORE1 places people for a living across eight verticals, and if you cut too deep and then need bodies fast, we are one of the companies you would call to fix it. That is a real conflict and you should read the rest of this knowing it. I would rather you never make the call. The section on what to do instead of a deep cut is the one I actually care about, and it costs us money when clients follow it. Our HR staffing team spends a lot of its week on exactly this conversation.

What a Boomerang Employee Actually Is

A boomerang employee is a person who worked for your organization, left, and was later rehired into the same company. The departure can be voluntary or involuntary. The distinction matters more than almost anyone writing about this admits, because the two versions behave nothing alike once they are back in the building.

The voluntary boomerang left for a better title, got bored, and came back with a wider view of the market. Usually a good bet. Take the meeting.

The involuntary boomerang is different. You ended their employment. They updated their resume, told their spouse, possibly moved a kid to a different school district, and then you called. Whatever they say in the offer conversation, a piece of the relationship got repriced permanently. Same badge. Different arrangement. Permanently.

Almost every glossary page on this topic collapses the two into one entry with a pros and cons list. That is the gap this piece is written into.

One Number That Should Change How You Model a Reduction

ADP Research tracks this in actual payroll records rather than survey responses, which is why I trust it more than most of what gets published in this space. Boomerangs are about 2 percent of all active employees. They are 31 percent of new hires on average going back to 2018.

By March 2025 that share hit 35 percent. A year earlier it was 31 percent. At the low point in March 2022, when everyone was quitting for a 20 percent bump and companies were hiring anybody with a pulse, it was 26 percent.

Now the number that should stop you. In the information sector, close to two thirds of new hires in March 2025 were returning employees. Two thirds. Double the rate from the year before, against a long-run sector average of 30 percent.

Read that as an operational signal rather than a hiring trend. Tech organizations are not rediscovering their alumni out of sentiment. They cut, discovered what the cut actually removed, and went back. The federal JOLTS report released August 4, 2026 puts layoffs and discharges at 1.8 million for June against 5.3 million hires, which tells you the churn engine is running hot in both directions at once. Companies are separating and rehiring simultaneously, frequently in the same function.

What It Costs to Buy Back the Person You Cut

Here is where the spreadsheet lies to you.

A reduction shows up in the model as immediate salary and burden savings. The cost of undoing it shows up eight months later in a different budget line, usually recruiting, usually owned by somebody who was not in the room for the original decision. The two numbers never get compared, which is precisely why the pattern repeats.

Some real figures to put against it.

The SHRM benchmarking numbers come from 2,371 members surveyed between January and March 2025. The Visier analysis sits on anonymized records for 2.4 million employees at 142 enterprises, and it contains the single most uncomfortable finding in this whole subject: the people most likely to get rehired after a layoff are the high performers and the managers. The exact population the cut was supposedly designed to protect.

Which means the reduction did not remove your weakest 8 percent. It removed a slice that included some of your best, and you will pay a premium to get a fraction of them back. The fraction is small. Roughly one in twenty.

The other nineteen are working for somebody else now. Several of them are working for the competitor who called them the week your announcement hit the trade press. I have made that call on behalf of clients. Takes about ten minutes. Costs nothing.

The Study Almost Nobody Quotes

Search this topic and you will get roughly forty articles telling you boomerangs are a bargain because they ramp faster and already know the culture. Neither claim is wrong. Neither is the whole picture either, and the reason you rarely see the other half is that the other half sits in a management journal instead of a vendor blog.

Researchers from Portland State, Purdue, Missouri, Iowa, and UT Rio Grande Valley ran the comparison properly. Their paper, published in the Journal of Management, tracked 1,318 boomerang managers against 20,850 external hires and 8,546 internal promotions at a large national retailer, with performance ratings from before and after the rehire.

Three findings, and they are not the ones on the vendor pages.

Boomerang managers performed about the same as internal and external hires in year one. No penalty there. But internal and external hires kept improving after that. The rehires mostly did not. Their performance after returning tended to hold flat at whatever it had been before they left. You did not get a better version of the person. You got the same version, at a higher salary, with a gap in the middle.

Then the part that should genuinely change your decision. Boomerangs turned over at a higher rate than either comparison group, and when they left the second time, the reasons looked a lot like the reasons they left the first time.

Nothing got fixed while they were gone. The commute was still the commute. The manager was still the manager. Nothing changed. Rehiring somebody into an unchanged situation and expecting a different outcome is a fairly expensive way to learn that.

I want to be careful here, because that study looked at retail management inside one organization. One employer, one sector. Not a universal law. But it is the most rigorous look anyone has published, and it points in the opposite direction from almost every article that will show up above this one in search results.

Why the Cut Goes Too Deep in the First Place

Nobody sets out to cut into muscle. It happens for three reasons, and I see all three constantly.

The first is that reductions get modeled at the layer of cost and executed at the layer of people. A finance team builds a target. An HR team allocates it across functions. A director picks names against a rubric that weights tenure, comp band, and last review cycle. Nowhere in that chain does anybody ask which four humans hold knowledge that exists in no document anywhere. That question has no field on the form.

The second is that the work does not disappear when the person does. Everyone assumes the remaining team absorbs it. For about eleven weeks, they do, and it looks like the model was right. Then the on-call rotation gets thin, two people quit voluntarily because their weeks became unbearable, and the savings evaporate into overtime, agency spend, and a search you did not budget for.

Third, and this one is newer, is that the stated reason for the cut is frequently not the actual reason. Forrester’s January 2026 forecast puts it about as bluntly as a research firm can. Forrester expects roughly 6 percent of US jobs, about 10.4 million roles, to be automated by 2030, with another 20 percent augmented rather than eliminated. And it expects more than half of layoffs attributed to AI to be reversed, on the grounds that plenty of companies announcing AI-driven cuts have no mature, vetted AI application actually ready to do the work.

Forrester has a name for that. AI washing. A financially motivated cut wearing a technology story, because a technology story reads better on an earnings call than “we over-hired in 2022.”

The reversal is the part to underline. More than half. Undone. That is a research firm telling you, in advance, that a majority of these decisions are going to be walked back at somebody’s expense.

Five Moves That Cost Less Than Cutting and Rebuying

None of these are clever. They are just the things that work, in rough order of how often they get skipped.

• Map the knowledge before you map the cost. Ask every manager one question: if this person resigned tomorrow with no notice, what breaks and how long until it is fixed? Anybody whose answer runs past three weeks comes off the list, or the knowledge gets documented first. Four hours per department, give or take. Nobody has ever regretted doing it. I have watched several regret skipping it.
• Cut shallower and bridge the gap with contract. If you need to hit a headcount number for a board deck, a lighter permanent reduction paired with contract staffing for the seasonal or project load hits the same run-rate target without severing institutional knowledge. The contractor leaves when the work leaves. That is the whole point of the model and it is chronically underused in exactly the moment it is most useful.
• 3.2 million people a month quit their jobs voluntarily, per JOLTS. Some of that is already happening in your building. You are just not counting it as a reduction. A hiring pause plus natural attrition over two quarters gets a surprising number of organizations most of the way to their number without a single involuntary separation, and without the announcement.
• Keep the list. The people you let go are the highest-signal candidate pool you will ever have access to, and most companies lose track of them within ninety days. Somebody should own that spreadsheet. Names, roles, what they knew, and honest notes about whether you would take them back. When the work comes back, and per Forrester it usually does, you are calling a warm list instead of starting a search from zero.
• Decide the comp position now, not later. The moment you decide to call somebody back, you have already lost most of your leverage. Work out in advance what a return is worth, including the premium, and compare it honestly against an external search. Our salary benchmark assistant will get you a defensible range in a few minutes, and if the return number is worse than a fresh search, that is useful to know before you dial rather than after.

Longer term, the fix is not tactical. It is planning. Our workforce planning guide walks through the modeling that keeps organizations out of this loop, and if AI is genuinely part of your headcount thesis rather than the cover story, we wrote about headcount models for AI transformation separately.

If You Already Cut Too Deep, Do It in This Order

Plenty of you found this page a quarter too late. Fair enough. There is a sequence that works better than panic, and it is not the one most leadership teams run.

Diagnose before you dial. Figure out whether you need that specific person or the capability they held. Those are different problems with different price tags, and leaders conflate them constantly. If it is the capability, the market has other people who have it. Some are cheaper. Most are hungrier.

Then, before anybody picks up a phone, deal with the reason they went. If the person quit ahead of the reduction rather than being caught in it, whatever pushed them out is still sitting exactly where they left it. The Journal of Management data is unambiguous on this point. Rehire into an unchanged situation and you are simply buying the same exit again, eighteen months out, at a higher price.

Make the call yourself. Not a recruiter. Not a coordinator. Definitely not a templated email from an address nobody replies to. If you sat in the room where their employment ended, you are the one who dials, and it converts at a rate that embarrasses every other approach.

Price it honestly and once. Lowball a boomerang and you confirm every private theory they have about how the company values them. They will take the offer if they need it, and they will leave the second something better appears, and you will have paid a search fee for a nine-month rental.

Last one, and it is the step people skip because it feels disloyal. Run an external search alongside the boomerang conversation. Not as a threat, and not as leverage in the negotiation. Purely as a price and quality check on a decision you are about to make with incomplete information. If the returning person really is the best option available, a parallel search proves it and you close without second-guessing yourself for a year. Our average time to hire on IT roles is 17 days, so the sanity check rarely costs real calendar time.

Things Leaders Ask Us After a Reduction

So what actually counts as a boomerang employee?

A boomerang employee is anyone your organization employed before, who left for any reason, and who later returns to your payroll. Voluntary exits and layoffs both count. The two behave very differently once they are back, so track them separately in your reporting.

Is rehiring somebody we laid off cheaper than hiring a stranger?

Cheaper on the invoice, not reliably cheaper on the P&L. You skip most onboarding and they ramp faster, which is real money. Against that, Visier finds returning employees take a 5 percent pay increase versus 2 percent for the people who stayed, and the Journal of Management research found rehires plateau where new hires keep improving. Run both numbers over three years rather than one quarter, and the gap narrows a lot. Sometimes to nothing.

How long do we have before the good ones stop taking our calls?

Just under six months is the average return window in Visier’s data, with most boomerangs back inside ten months. In practice your window on a strong senior engineer is far shorter than that. Two to three weeks in a competitive stack. They had a recruiter in their inbox before the severance paperwork cleared, and the good ones do not sit unemployed.

Do rehires actually stay?

Depends entirely on whether the reason they left the first time still exists. The Journal of Management study of 1,318 boomerang managers found they turned over at a higher rate than internal or external hires, and their second departures looked like their first. If you changed the manager, the scope, or the comp, the odds improve. If you changed nothing, you bought a delay.

Should we tell people we might bring them back?

Short answer: only if you mean it. A vague “we hope to reconnect when things improve” costs you nothing and buys you nothing, and everybody in the room knows what it is. What works is specific. Name the quarter, name the conditions, and put someone’s actual name on the follow-up. We have seen clients successfully rehire eighteen months out on the strength of one honest exit conversation.

Does using a staffing firm to rehire our own alumni make sense?

Frequently no, and I will tell you that on the intake call. If you have the name and the relationship, calling them yourself is faster and free. Where a third party genuinely helps is when the relationship is damaged, when you need a market read on whether the return price is fair, or when you need three alternatives alongside the boomerang so the decision is a real comparison. That is worth a fee. Simply passing along a phone number is not.

The Part Worth Remembering

A layoff is a bet that the work shrank as much as the headcount did. Sometimes it did. Often it did not. When it did not, the correction arrives about six months later, costs somewhere between $5,475 and $35,879 per seat before the pay premium, and hands you a person who, per the best available research, will probably perform exactly as they did before and leave sooner than the alternatives.

Two thirds of information-sector new hires being boomerangs is not a heartwarming statistic about loyalty. It is an industry publicly grading its own workforce planning.

KORE1 has been placing technology and professional talent since 2005, across 30-plus US metros, with recruiters who average more than 15 years in the market and a 92 percent 12-month retention rate on the people we place. A fair amount of that work is helping companies rebuild after a cut went further than intended. More of it lately is helping them size the cut correctly the first time, which is cheaper for everyone including us. If you are modeling a reduction this quarter and want an outside read before the names are final, talk to one of our recruiters. That conversation is free and it is the one we would rather have.