What Counts as Job Hopping? Less Than Almost Anyone Thinks

There's an email every recruiter knows. It arrives within a day of a candidate's profile landing, and it goes something like: they're only a few months into their current role, they didn't stay long at the one before – they've jumped every time, so they'll jump from us too. Deal breaker.

A handful of dates on a CV, read as a prediction. We've received many versions of that email, and whenever we've been able to check the prediction against what we actually hold on the candidate, it has come apart in the same few ways. One of the "jumps" turns out to be a layoff that took the whole team with it. Another is the exact career move the same hiring manager praised at kick-off as the mark of pedigree in their field. And the short current stint comes with a reason the candidate volunteered, on record, in interview – sometimes one that makes them more suited to the role, not less.

The dates are always real. What they're taken to mean very often isn't.

We went looking at how often that happens – across every hiring conversation we've recorded, and across 18,000 of the candidate work histories on our platform. The answer changed how we think tenure should be read, and it starts with an uncomfortable finding: nobody can tell you what "too short" actually means.

Nobody agrees what "too short" means

We run recorded calls with every hiring team we work with, so we went back through them. Twenty-five different organisations have raised candidate tenure with us – as a requirement at kick-off, or as a reaction to a specific candidate. Between them they named sixteen different thresholds for what counts as too short: six months, nine months, a year, two years, thirty months, three years – and ten more in between.

The reasons behind the thresholds disagreed even more. One sales founder uses tenure as proof of performance: survive several years in an enterprise seat and "at least you haven't been fired". A CEO reads it as character – people "hopping from job to job" seem "unsure that they have what it takes to commit to one thing". A hiring manager priced it as onboarding cost: "so much investment in somebody to get them trained up… you don't want them to leave". A recruiter told us she flags short stints because "you know your client's going to push back" – screening on someone else's anticipated objection. And several, pressed for a reason, offered none at all: "I think every hiring manager cares about those things."

Here's the pattern that convinced us this is folklore rather than judgement: almost everyone cancelled their own rule within a minute of stating it. "Sometimes I feel like it's unfair to assume that you know why that is because they can always be reasons." "Red flags can be explained… the red flag should not be a rejection." Even the hardest line we heard – three years per stint, minimum – was waived in the same breath, in the same call, for a candidate whose short stints were a layoff and a poaching: "that trajectory is nice."

A rule this widely held, this differently defined, and this quickly abandoned under its holder's own scrutiny is a habit, and habits deserve to be tested against data.

Most job-hopping is a counting error

Before testing whether short tenure predicts anything, you have to measure it properly. Two counting mistakes distort the numbers.

Roughly a quarter of the "job-hopping" visible on a CV is not jobs. Internships, co-ops, fixed-term contracts, freelance projects and part-time work are short by design. Counted naively, 54% of the candidates in the sample have held a job for under a year. Strip out everything that was never a permanent full-time role, and it falls to 42% – and the median completed stint moves from 17 months to 21 months. A four-month co-op term is a credential, and half the world's screening processes are reading it as a defect.

The second mistake is one people rarely make and rules make constantly. A written tenure rule, or a tenure statistic, must count employers, never positions – merging every promotion and internal transfer before it judges anything. Nobody looking at a profile mistakes a promotion for a job change. But the moment the bar is written down – "two years minimum per stint" – and applied mechanically, every internal move starts counting against the candidate, because LinkedIn lists each one as a separate entry. We've watched a client's written tenure rule fail exactly this way, rejecting people whose "short stints" were promotions inside one company, until it was patched. The distortion is not small: across 18,000 of our candidate histories, the median candidate is 23 months into their current position and 36 months into their current employer – so a rule or statistic built on the per-position number runs 13 months low at the median, and lands hardest on exactly the people who got promoted.

Get the counting right and a different picture appears. The median candidate in our data is nearly three years into their current employer. That's the population hiring managers describe to us as a generation that never stays anywhere. The perception gap comes from counting the CV the way LinkedIn formats it rather than the way careers actually run – a cousin of the problem we wrote about with years-of-experience filters: the convenient number on the page and the thing you care about are different things.

The people you hire have short stints too

Now the question that actually matters. Employers screen out short tenure to avoid a bad outcome. So do the candidates who succeed – the ones who get interviewed, shortlisted, and hired – have cleaner dates than everyone else?

In our data: no. We measured every candidate the same way (permanent full-time roles only, counted by employer) at every stage of the funnel. 32% of all candidates have a completed sub-year stint sitting in the same career as a three-year-plus stint. Among the candidates who reached a recorded interview with us: 32%. Among the candidates whose profiles we put in front of clients: statistically indistinguishable from the baseline. The share barely moves at any stage. Short stints and long stints live in the same careers, at the same rate, no matter how far the person goes – strong people leave quickly when something isn't working, and stay for years when it is.

The hires themselves make the point better than the funnel does. Of the placements we've made, the median candidate was 14 months into their current role when we first spoke to them – and two were hired at four and six months in. One of our most successful placements had – counted properly, by employer – a median completed stint of seven and a half months: four of her six stints ran under a year, alongside one three-year spell. Her employer hired her with all of that on the table, and she's still in the seat. (This is a small sample – we publish it as evidence about how the filter behaves, not as a statistical law.)

We've also watched the filter fail in real time. On one search, a client's tenure requirement – one we had helped write – turned out to exclude two people the client had already hired before we arrived. When we showed them that, they rewrote the rule themselves. We'll come back to what they replaced it with, because their replacement is the best definition of job-hopping we've seen.

Short stints are mostly the employer's story

The job-hopper reading assumes a short stint records the candidate's decision. Much of the time, it records the employer's.

In the year to June 2026, roughly one in three US job separations was a layoff or discharge – employer-initiated, not chosen. 4.1 million American workers were displaced from jobs they had held under three years in 2023–25 alone. And employers themselves are less stable than the loyalty conversation assumes: about one in five new establishments doesn't survive its first year, and roughly half are gone within five. A candidate can collect three short stints in a row without making a single bad – or voluntary – decision.

Our own interview recordings say the same thing. When candidates explain a short stint on our calls, the reasons that come up over and over are layoffs, restructures, leadership changes and companies that ran out of road. One candidate had spent fifteen years at one employer before a restructure ended it. Another lost two jobs to team-wide cuts – four hundred people in one of them. A third told us, unprompted: "I like to stay for a long time. I don't like to jump around… it was very unforeseen on my end." In our recorded interviews this year, no candidate has ever explained a short stint as boredom.

The generational version of the worry doesn't survive the data either. US median tenure for 25-to-34-year-olds was 2.7 years in January 2024 – against 3.0 years in 1983 and 2.6 in 2000. Four decades, essentially flat. The generation now doing the hiring averaged 12.9 jobs between ages 18 and 58, and ended 57% of the jobs they started between 25 and 34 within two years. Early-career movement has been how careers work for as long as it has been measured: the classic study of young workers found a typical worker holds about seven jobs in their first ten years, with job changes driving at least a third of early wage growth.

What the job-hopping worry gets right

None of this means the signal is empty, and pretending otherwise would be dishonest. The strongest evidence on the other side is real. A field experiment in Switzerland sent employers matched CVs with identical total experience – one employer versus four – and the one-employer version got roughly 40% more callbacks. The penalty exists, and employers act on it. Longitudinal research also finds that workers with a history of frequent moves are more likely to move again – but that effect is concentrated in voluntary quits and disappears for layoffs, which is exactly why the reason behind each move carries the information the dates can't.

There's a sharper version of the worry worth naming: layoffs aren't lottery draws. When a company cuts selectively, who stays is partly a performance decision – labour economists have long documented that the market treats a selective layoff as a worse signal than a plant closing, where nobody was chosen. But notice where the information actually lives: in the difference between "I was picked" and "the whole team went" – a distinction the CV cannot show and one interview question can. And for short stints it runs weakest of all, because layoffs fall hardest on the most recently hired. Where a layoff really was about the person, the definition below already catches it: an exit that was about performance is a performance exit, whatever the paperwork called it.

Two more findings sharpen where the signal genuinely lives. Psychologists studying why people hop find two distinct motives with opposite profiles: escape-driven hopping tracks with impulsivity, while advancement-driven hopping tracks with persistence and deliberate career management. The CV shows the behaviour and hides the motive. And the norms are role-specific: a US field experiment across 3,878 job postings found the job-hopper callback penalty holds in most occupations and reverses in some technical ones. In our own data, median current-employer tenure ranges from 16 months to 75 months depending on the role we're searching for – a 4.7x spread. Any single threshold applied across roles is wrong for most of them.

The definition that survives

Remember the client whose tenure rule excluded two of their own best hires? Here is what they replaced it with, and it has held up on every search since: a pattern of short stints ending in terminations or performance exits is the red flag – moving for money or opportunity is fine, and the reason for each move is the thing to probe. It reads like a rule about time. It is really a rule about reasons.

Put together with the counting rules, that gives a definition employers can actually write down. Job-hopping, properly defined, is three or more consecutive short stints in permanent full-time roles, counted by employer, that were voluntary, gained the candidate nothing, and can't be explained when asked. Everything else that looks like it, isn't:

• Internships, co-ops, contracts and interim work – short by design. Exclude them from the pattern entirely.
• Layoffs, restructures, acquisitions and company failures – the employer's decision, recorded on the candidate's record.
• Being recruited into a bigger role – other employers competing for someone is evidence for them, whatever it does to the dates.
• One short stint, even a messy one – a single bad fit is base-rate behaviour. I have two five-month stints on my own CV. I left both because the roles were wrong, and I've stayed three-plus years everywhere else.
• Early-career movement – in the first five years, moving is how people find the work they'll later stay in.

What remains after those exclusions is rare. Research on the dispositional "hobo syndrome" – people who move compulsively, job after job – puts them at under 5% of the workforce. Meanwhile 38% of employed 25-to-34-year-olds are under two years into their current job at any moment. A blanket tenure screen fires on a third of the market to catch a twentieth – and the twentieth it's aimed at is findable by a much better tool anyway.

That tool is the conversation. Ask about each short stint, specifically: what happened, whose decision was it, what did the move gain you? Candidates with real reasons answer in seconds – often, in our experience, before you finish the question. The rare compulsive mover produces vague circumstances and no gains. The one candidate we've watched a client cut on tenure mid-process was cut because their answer fell apart live, and that was the right call – made on the reason, where the information actually is, and consistent with how a structured interview should test any hypothesis a CV raises. If the worry matters enough to reject on, it matters enough to write into the role's requirements and apply to every candidate the same way – a twenty-minute conversation is a cheap price for not losing a third of your qualified market to folklore.

Find the talent you're looking for

lemonly delivers qualified candidates on-demand. Find out how.

Book A Demo

Pick a time for your demo on the next page.

Key takeaways

• Twenty-five employers gave us sixteen different definitions of "too short tenure" – and almost all of them talked themselves out of their own rule within a minute. The threshold is folklore, and most versions of it flag the median worker.
• About a quarter of apparent job-hopping is not jobs – internships, co-ops and contract work. Strip it and the sub-year-stint share drops from 54% to 42%.
• Candidates who progress furthest look no more "stable" than anyone else: roughly a third at every funnel stage have a sub-year stint next to a 3-year-plus stint. Strong people leave fast when it's not working.
• About one in three separations is the employer's decision, and the compulsive mover the screen is aimed at is under 5% of the workforce. Define job-hopping as a pattern of unexplainable, gain-free voluntary exits – then test it in the interview, where the reasons live, and apply the same written rule to everyone.
• And if the rule is written down or automated: count employers, never positions – applied mechanically, per-position counting reads promotions as short stints and understates the median candidate's current tenure by 13 months (23 vs 36 months in our data).