You have heard the number. "80% of people who accept a counteroffer are gone within six months." Sometimes it is 90% within a year. Every agency deck, every "never accept a counteroffer" LinkedIn post, every candidate-closing script leans on it. We went looking for the primary source and found nothing: no survey, no methodology, no year, no named researcher. Just the same round number quoted by people whose business improves when candidates believe it.
Here is the strange part. The folklore is unnecessary, because the numbers that do have sources tell a bad-enough story on their own. This is the full audit: what is verified, what is recycled, and what each side of the desk should actually do with it.
The verified counteroffer statistics
Every number in this table traces to a named study with a methodology and a date.
| Statistic | Number | Source |
|---|---|---|
| Managers who make counteroffers to employees trying to leave | 58% | Robert Half survey of 5,500 hiring managers, 2018 |
| Average time an employee stays after accepting a counteroffer | 1.7 years | Robert Half, 2018 (as reported by senior managers) |
| Workers who have received a counteroffer at some point in their career | 57% | Robert Half UK, 2022 |
| Workers who accepted the counteroffer the last time they got one | 53% | Robert Half UK, 2022 |
| Counteroffer acceptors gone within six months | 34% (47% for under-35s) | Robert Half UK, 2022 |
| Counteroffer acceptors gone within a year | 74% | Robert Half UK, 2022 |
| Acceptors out the door within twelve months | ~50% | Research cited by Harvard Business Review, 2016 |
Primary write-ups: SHRM's coverage of the Robert Half manager survey, HR Dive's summary of the same data, and the Robert Half UK study published via theHRDirector in March 2022.
Read the table twice and a shape emerges. Counteroffers are common: more than half of managers make them, more than half of workers have received one, and more than half of recipients say yes. And they buy time rather than loyalty: depending on the study, somewhere between half and three-quarters of acceptors leave within a year anyway.
Where the 80/90 folklore comes from
Chase the "80% gone in six months, 90% in a year" claim upstream and the trail runs through recruiting agency blogs, a risk-management association's career page, and posts citing "national statistics" with no nation, no statistic, and no year. The most commonly named origin, a UK recruiter's survey, is itself a small self-run poll rather than published research, and the numbers mutate between retellings: 80% within six months in one version, 80% within a year in another, 89% and 93% in others.
That pattern is the tell. Real survey findings keep their shape because there is a document to check. Folklore drifts because there is not. A useful working rule for any recruiting statistic: if you cannot find the study, the number is a sales asset, not a fact. It is the same discipline we apply to layoff coverage, where the announced reason and the data routinely disagree.
None of this means the myth points the wrong way. The verified numbers agree that counteroffer acceptors churn fast; they just put the one-year loss at 50 to 74 percent instead of 90, and the six-month loss at about a third instead of 80 percent. If you are a recruiter, quoting the sourced version makes you more credible, not less persuasive. The candidate who catches you repeating an untraceable stat discounts everything else you said.
Why counteroffers fail (when they fail)
The mechanism is not mysterious. A counteroffer usually raises the salary and changes nothing else, while pay is rarely the only reason someone interviewed elsewhere. The manager relationship, the growth ceiling, the on-call load, the strategy doubts all survive the raise. Robert Half's own senior district director describes counteroffers as "often only a temporary fix" and advises employers to prepare for the employee to leave within six months regardless.
There is also a trust tax that is hard to measure and easy to observe: the employee has now shown they will look, and the employer has now shown what the work was worth all along. Both sides remember. The under-35 number (47% gone within six months) suggests younger employees treat an accepted counteroffer even more explicitly as a bridge to the next move.
What recruiters should do with the real numbers
The counteroffer is not an anomaly at the end of your process; at current acceptance rates it is the expected final competitor on more than half of your placements. Three moves follow from the data.
Surface the leaving reason at the start, not the end. If money is the only stated motivation, a counteroffer will beat you, because the current employer can always match faster than you can close. The intake conversation logic applies to candidates too: the reasons that survive a raise (manager, ceiling, mission) are the ones worth building your close on. The same early questions that predict offer rejections predict counteroffer acceptance.
Rehearse the counter before resignation day. A candidate who has already heard "your employer will likely offer you more money, here is what will and won't change" is inoculated; one who hears it for the first time from their boss is not. Use the sourced numbers in that conversation: 53% of people accept, and half to three-quarters of them are searching again within a year.
Anchor the offer to market, not to the current salary. Counteroffers win when the new offer merely matches what the employer can trivially raise. When the offer sits at verified market rate for the role and metro, matching it costs the incumbent employer real money, and the conversation shifts from "we'll match" to "we can't". Live pay benchmarks by role, seniority, and geography are what Glozo Intelligence is for, and checking them before the offer stage is cheaper than losing a placement after it. Losing one means eating the search cost twice; the fee math is in our recruiting fees breakdown.
What employers should do with them
The same data reads differently from the hiring side. A counteroffer that keeps an employee for 1.7 more years is not automatically a failure: if the alternative is an empty senior seat, a search fee, and a six-month ramp, buying 20 months at the cost of a raise can be rational. The mistake is treating the accepted counteroffer as a resolution rather than a deadline.
The honest playbook: counter only when the person is expensive to replace and the fix addresses their actual reason for looking; start succession planning the day they accept; and treat the episode as free market data. An employee fielding outside offers at 20% above their current pay is telling you your pay bands drifted below market, and they are rarely the only one affected. Checking your bands against live market data once a quarter costs less than learning about the drift one resignation at a time.