# The Tenure-Pattern Read: Advance, Probe, or Screen Out

*You will score a candidate's work history against the right tenure baseline and output a defensible advance, probe, or screen-out verdict with one resolving question per stint.*

- Canonical URL: https://www.refolk.ai/guides/tenure-pattern-read-2
- Pillar: Recruiting and sourcing
- Format: Framework
- Published: 2026-10-07
- Last reviewed: 2026-10-07
- Reading time: 17 min

A candidate reaches your desk with three or four short stints on their profile and a hiring manager waiting on your read. Before you advance, you have to decide whether that work-history pattern is a real risk or a non-issue, and you have to be able to defend the call. This guide gives in-house recruiters, sourcers, and founders doing their own hiring a scored rubric with stated tenure thresholds, legitimate-cause carve-outs, and sector-and-age adjustments, so two reviewers grade the same history the same way and produce an advance, probe, or screen-out verdict plus the one question that resolves each ambiguous stint.

The top search results argue whether job hopping is still bad and leave you with an opinion. An opinion does not survive contact with a hiring manager who asks why. A rubric does.

## What this read decides, and why gut feel fails it

This read turns a set of dates into a repeatable verdict: advance the candidate, probe a specific stint on the screen call, or screen out. The failure it fixes is two experienced reviewers looking at the same timeline and disagreeing, because each is scoring against a private baseline and a private threshold.

The disagreement is not hypothetical. The surveys split hard. A LinkedIn survey of 1,024 hiring managers found 37 percent said frequent job changes might stop them pursuing a candidate. A survey of 200 recruiters found 52 percent consider job hopping a red flag and 84 percent have turned down applicants over it. Resume Genius, by contrast, reported that 50 percent of employers do not consider job hopping a red flag at all. When half the market flags a behaviour and half does not, your personal instinct is a coin toss dressed as judgement.

**37-52% - Share of surveyed recruiters who treat job hopping as a red flag**

The same behaviour flags for roughly half and passes for the other half, which is why a shared threshold beats instinct.

The rubric closes that gap by making four things explicit: the baseline you compare against, the pattern count that triggers concern, the carve-outs that excuse a short stint, and the trajectory across moves. Score all four and the verdict writes itself. Skip any one and you are back to arguing.

## The baseline choice decides half the verdict

Pick the wrong reference distribution and you will over-flag or under-flag before you score a single stint. The national median tenure was 3.9 years in January 2024, the lowest since 2002, but that figure is a blend across every age and industry and it fits almost no individual candidate.

Two adjustments matter most. The first is age. Younger workers churn more by default, so a short-stint timeline on an early-career profile is usually normal, not a warning.

| Age band | Median tenure (years) |
| --- | --- |
| 20-24 | 1.4 |
| 25-34 | 2.7 |
| 35-44 | 4.6 |
| 45-54 | 7.0 |
| 55-64 | 9.6 |

A 24-year-old with two 15-month roles is sitting on top of their band median, not under it. The 55-64 median of 9.6 years is more than three times the 25-34 median of 2.7 years. Grade age-blind and you penalise the young for behaving like everyone their age.

The second adjustment is sector. The spread here is wide enough to flip a verdict on its own.

| Sector | Median tenure (years) |
| --- | --- |
| Public sector | 6.2 |
| Mining, quarrying, oil & gas | 5.7 |
| Manufacturing | 4.9 |
| Private sector (all) | 3.5 |
| Leisure and hospitality | 2.1 |

A two-year stint is a flag against the 6.2-year public-sector median and completely normal against the 2.1-year leisure-and-hospitality median. Tech runs below even these: one compilation of roughly 103,000 developers found 45 percent of software developers have an average tenure of just one to two years and 69 percent stay under two years, with big-tech tenure often pegged near two years. In tech, marketing, and startups, changing jobs roughly every 18 months is frequently viewed as standard, while law, finance, and manufacturing hold the line at three or more years.

> **Watch out:** The most common scoring error
>
> Grading a developer's two-year stints against the 3.9-year national median makes a normal career look disloyal. Always re-score against the sector and age baseline before you flag anything.

Seniority flexes the read too. For entry-level professionals, frequent job changes are widely accepted; for mid-to-senior managers, frequent changes raise real questions about the ability to execute long-term strategy, because the job itself is measured over multi-year horizons.

## The pattern rule: count stints, not a stint

One short stint is a data point. A pattern is three or more consecutive sub-two-year roles inside a five-year window. That is the only threshold you can defend, because any single short stint scored as disqualifying contradicts every source that defines the behaviour.

The sources converge on a pattern, not a cutoff. One glossary frames job hopping as changing employers frequently with stints typically under two years, by choice rather than through layoffs or contract endings, and notes that one short stint is a data point while three in a row is a pattern recruiters will ask about. Another defines it as holding multiple consecutive jobs each lasting 12 to 18 months and stresses that recruiters look for the pattern over a three- to five-year window. A recruitment glossary puts it at an average tenure below 18 to 24 months across two or more consecutive roles. On the acceptable floor, a Randstad director said a minimum of 18 months at a job is acceptable, though a tenure of three to five years is the stronger signal, with the sweet spot often cited as two to three years.

> **Rule:** Pattern before penalty
>
> Do not flag a candidate for tenure until you can point to three or more consecutive roles each under about two years within a single five-year window. Fewer than that is a data point, not a pattern.

Apply this literally. Lay out the timeline, mark each stint short or normal against the baseline you just set, and look for three consecutive shorts packed into five years. A profile with one 10-month role surrounded by multi-year tenures has no pattern, whatever your gut says about the 10 months.

#### From dates to a defensible verdict

1. **Baseline** - Look up the sector and age median for this candidate
2. **Map** - Tag every stint short or normal against that baseline
3. **Count** - Test for three-plus consecutive shorts in five years
4. **Carve out** - Label each short stint voluntary or structural
5. **Trajectory** - Grade title, scope, and comp direction across moves
6. **Verdict** - Advance, probe, or screen out with a written rationale

*Each stage narrows the judgement so the final call cites a baseline, a count, a carve-out status, and a trajectory.*

## The carve-outs: when a short stint is structural, not a choice

Some short stints are not the candidate's doing, and three of these are publicly verifiable. The carve-outs are contract or interim roles, mass layoffs, acquisitions, and company closures. The definition of job hopping itself excludes stints that ended by layoff or contract expiry rather than by choice, so applying these carve-outs is not generosity, it is grading the behaviour the sources actually name.

Contract and interim work is the largest carve-out and the easiest to miss, because the dates look identical to churn. The category is far bigger than most reviewers assume, and it is uneven across markets.

| Market | Contract/interim-titled profiles | Share of US |
| --- | --- | --- |
| United States | 333,355 | 1.00x |
| United Kingdom | 73,749 | 0.22x |
| US-to-UK ratio | 4.52x | - |

In Refolk's index of professional profiles there are 333,355 US profiles and 73,749 UK profiles carrying contract, interim, or consultant-style titles, which means interim work is 4.52 times more common in the US than the UK. A reviewer screening cross-border candidates should expect a string of short engagements far more often in US histories and should read the title, not just the dates. A fractional CFO with five 14-month engagements is not a flight risk; that is the job.

Layoffs are now cheaply falsifiable, which changes how much benefit of the doubt an unverified claim deserves. The WARN Act requires businesses to give 60 days of notice for plant closures or mass layoffs, and states publish those notices. One aggregator holds more than 85,100 WARN notices covering 9.0 million workers, standardising filings from 50 states and territories into searchable datasets.

**85,100+ - WARN layoff notices searchable by company, covering 9.0M workers**

A claimed mass-layoff exit with no matching filing is now a check-worthy signal, not an automatic carve-out.

Acquisitions and closures are checkable too. Crunchbase records organization status as Active, Closed, Operating, Acquired, or IPO, so a "company was acquired" claim has a public answer. If a former employer was acquired or merged, the acquiring company's HR may also hold the inherited employment records, which matters at background stage.

One honest note on sequence: practitioners disagree on whether to carve out causes before counting the pattern or after. Carving out first changes the count, because a layoff stint removed from the sequence can break a three-in-a-row chain. I carve out first, because a structural exit was never the candidate's behaviour and should not inflate the pattern. State which order you used in the rationale so a second reviewer can reproduce it.

> **Watch out:** Do not trust resume-stated causes
>
> "Company closed" or "role eliminated" on a resume is a claim, not a fact. Verify structural causes at reference or background stage through WARN filings, Crunchbase status, or payroll-provider records before you let them excuse a stint.

## Progression versus churn: grading the trajectory

Count is not the whole story; direction is. Two candidates with identical stint lengths can be moving up or treading water, and the difference decides between a probe and a screen-out. Grade trajectory on three markers: title tier, scope, and compensation direction.

An upward move is a promotion to a higher-level role, usually with more responsibility, leadership duties, and compensation. A lateral move is a shift to a different role at the same level. The warning sign is not any single lateral move, it is repeated sideways motion with no upward movement over a long window, which raises real questions about career trajectory.

But do not reflexively mark lateral as stagnation. A Management Science study found lateral mobility has an increasing effect on subsequent advancement, most significant in years three and four after the move, representing around one fifth of a pay grade. A deliberate skill-broadening move can be the setup for a faster climb, not evidence of drift. Trace the direction across the whole history rather than scoring each move in isolation.

#### Pattern count against trajectory

Horizontal axis runs from No pattern (fewer than 3 shorts) to Clear pattern (3+ shorts in 5 years). Vertical axis runs from Flat or declining trajectory to Upward trajectory.

| Quadrant | What it means |
| --- | --- |
| Watch, likely advance | Short tenures but rising; confirm the moves were chosen |
| Advance | Normal tenure and climbing; no tenure concern to raise |
| Screen out candidates | Pattern with no progression; the clearest decline case |
| Probe | Pattern but rising; carve-outs and intent decide the call |

*Where a candidate lands on these two axes drives the advance, probe, or screen-out call.*

> A pattern with no progression is the clearest screen-out; a pattern with a rising trajectory is a probe, not a pass.

## The scoring procedure

Work the candidate through these seven steps in order. The whole read takes under an hour and produces a written rationale a hiring manager can read in a minute.

#### Score a tenure pattern end to end

1. **Set the sector-and-age baseline first** - Pull the candidate's industry and age band and look up the matching median tenure so you compare to the right norm, not the 3.9-year national figure. Write the expected tenure down before scoring any stint.
2. **Map every stint with dates** - List each role, its length in months, and its title in a clean timeline. Tag each stint short, under about 18 to 24 months, or normal.
3. **Count the pattern, not the stint** - Apply the three-in-a-row-within-five-years heuristic to decide whether a pattern exists at all. One short stint is a data point, not a pattern.
4. **Apply legitimate-cause carve-outs** - For each short stint, check for contract or interim framing, a WARN filing, an acquisition, or a closure, and label it voluntary or structural. State whether you carved out before or after counting.
5. **Score progression versus churn** - Check whether title, scope, and compensation rose, stayed flat, or regressed across moves, and tag the trajectory upward, lateral, or declining.
6. **Produce the verdict** - Combine baseline, pattern count, carve-outs, and trajectory into advance, probe, or screen out, with a rationale that cites each of the four inputs.
7. **Write one resolving question per ambiguous stint** - Draft a screen-call question that would resolve each unexplained short stint, and verify any claimed structural cause at reference or background stage.

The hardest part is the baseline lookup and the carve-out search, because they require pulling the right norms and checking sources outside the resume. Finding the comparable population by sector, stage, and geography, and separating contract-titled profiles from permanent ones, is exactly the kind of structured people search that [Refolk](/) turns into one plain-English request instead of a day of manual filtering.

I ran this search: `Fractional or interim CFOs in the US who have held three or more contract finance roles` - [see the full result list](https://www.refolk.ai/s/3cwtfpg9v3).

*Returns a list of profiles whose repeated short finance engagements are explicitly contract work, so you can see the carve-out population rather than mistaking it for churn.*

Once you can summon the contract-titled population on demand, the carve-out stops being a judgement call and becomes a lookup. The same holds for the opposite query when you want proof of stability: you can ask for engineers who stayed three or more years at each of their last two companies and compare the candidate against real peers rather than a national average.

## How this read goes wrong

The failure modes below are the false positives that make a tenure read indefensible. Each has a check that catches it. This is the part of the rubric worth memorising, because a wrong screen-out is invisible: the candidate just never advances, and nobody audits the miss.

| Failure mode | False positive it creates | The check |
| --- | --- | --- |
| Scoring against the 3.9-year national median | A developer with 2-year stints looks disloyal | Re-score against the sector baseline |
| Counting one short stint as a pattern | Flagging a single 10-month role | Require 3 consecutive shorts in 5 years |
| Treating contract work as churn | An interim string read as instability | Confirm the title and role type, not dates |
| Ignoring structural exits | A layoff or acquisition counted against the candidate | Search WARN and Crunchbase status |
| Reading lateral moves as stagnation | A skill-broadening move graded as going nowhere | Trace title, scope, and comp direction |

Two more deserve explicit mention. The first is over-weighting a single survey: asserting that recruiters reject job hoppers from the 52 percent figure while ignoring the 50 percent who do not flag it. Cite the range, 37 to 52 percent, or you are quoting a statistic to win an argument the data has not settled. The second is age-blind grading. Penalising an early-career candidate for short stints that are normal for their band ignores that the 20-24 median is 1.4 years. A junior candidate sitting at their band median is behaving exactly as expected.

The deepest trap is trusting resume-stated causes. "Company closed" taken at face value is a carve-out you have not earned. Verify at background stage through a WARN filing, Crunchbase status, or the acquiring company's inherited payroll records. With 85,100-plus WARN notices searchable by company, a mass-layoff claim with no matching filing should lose the benefit of the doubt, not keep it.

## The screen-call questions and the verdict record

Every ambiguous stint gets one resolving question, written before the call so the conversation is targeted rather than fishing. The question should be answerable in a sentence and should surface the carve-out or confirm its absence. Resolve the ambiguity on the call and verify the structural claims later.

**Screen-call questions by stint type**

```
Short stint, no stated reason: "Walk me through the move out of [company] after [N] months. What drove the timing?"
Claimed layoff or closure: "Was that a role-level cut or a broader layoff? Roughly how many people went?"
Claimed acquisition: "Who acquired [company], and did your role change or end at close?"
Contract or interim stint: "Was that an engagement with a defined end date, or a permanent role you chose to leave?"
Repeated lateral moves: "Those moves kept you at a similar level. What were you building toward across them?"
```

*Pick the line that matches the ambiguity you flagged in step 7. Keep each to one question and listen for a checkable fact.*

The verdict record is what you hand the hiring manager. It cites all four inputs so the decision is reproducible by anyone who reads it.

**Tenure verdict record**

```
Candidate / role under consideration:
Baseline used (sector + age median):
Stints tagged short: [list] / total stints:
Pattern present (3+ shorts in 5 yrs)? Yes / No:
Carve-outs applied (contract / layoff / acquisition / closure):
Trajectory (upward / lateral / declining):
Verdict: Advance / Probe / Screen out
Rationale (one or two sentences tying the four inputs together):
Open questions to resolve on screen call:
```

*Fill one per candidate. The rationale must reference baseline, pattern count, carve-outs, and trajectory.*

Before you call the read done, run this check.

#### Before you record the verdict

- [ ] The baseline is the candidate's sector and age median, not the 3.9-year national figure.
- [ ] Every stint is tagged short or normal against that baseline with its length in months.
- [ ] The pattern test required 3+ consecutive shorts in a 5-year window.
- [ ] Each short stint is labelled voluntary or structural, and the carve-out sequence is stated.
- [ ] Structural claims are marked for verification at background stage, not accepted from the resume.
- [ ] Trajectory is graded on title, scope, and comp direction, not title-level sameness.
- [ ] The verdict cites all four inputs and names the open questions for the screen call.

## Keeping the baselines current

The method is evergreen; the numbers inside it are not. The BLS Employee Tenure release updates on a roughly two-year cycle, so the medians in the age and sector tables will drift. Re-pull them from the BLS release before each hiring cycle rather than trusting a figure you memorised, and watch the national median in particular, which fell from 4.1 years to 3.9 years and sits at its lowest since 2002. Sector-level tech tenure is reported by third parties on their own schedules, so treat the 69-percent-under-two-years figure as a directional truth about the field rather than a fixed constant.

The carve-out sources change in the other direction, toward being more complete. WARN aggregators add notices continuously, so a layoff claim that was unverifiable last quarter may be searchable now. Re-check before you give a structural exit the benefit of the doubt. The contract-work population in Refolk's index also shifts as the labour market moves between permanent and interim work, and the US-to-UK gap of 4.52x is the kind of ratio worth re-reading when you start screening a new market. The rubric holds; refresh its inputs and it keeps producing verdicts two reviewers can agree on.

## Frequently asked questions

### How many short stints is too many on a resume?

The defensible trigger is three or more consecutive roles each under about two years within a five-year window, not a single short stint. Multiple independent sources converge on this pattern rule. One short stint is a data point recruiters may ask about; three in a row is the pattern worth flagging. Below that threshold, scoring a candidate as a job hopper is hard to defend to a hiring manager.

### Is short tenure always a red flag?

No. Short tenure only reads as a flag against the right baseline. The national median is 3.9 years, but 69 percent of software developers stay under two years, and the median for workers aged 20 to 24 is just 1.4 years. In tech, marketing, and startups, changing jobs roughly every 18 months is often treated as standard, while law, finance, and manufacturing expect three or more years.

### How do I verify a candidate's claim that they were laid off?

Check the state WARN database for a matching mass-layoff or closure filing, since the WARN Act requires 60 days of notice and states publish these notices. One aggregator holds over 85,100 notices covering 9.0 million workers across 50 states. For acquisitions and closures, Crunchbase records organization status as Active, Closed, Acquired, or IPO. Verify at background or reference stage rather than taking the resume at face value.

### Should I treat contract and interim roles as job hopping?

No. Contract, interim, and consultant roles are a known category of short engagement, not instability. In Refolk's index there are 333,355 US profiles and 73,749 UK profiles carrying contract-style titles, which is 4.52 times more in the US than the UK. Look for the contract framing in the title and role type, then exclude those stints from the pattern count.

### Do lateral moves count against a candidate's trajectory?

Not automatically. A Management Science study found lateral mobility raised subsequent advancement most in years three and four after the move, worth about one fifth of a pay grade. Grade trajectory on title, scope, and compensation direction across moves rather than on title-level sameness. Repeated sideways motion with no upward movement over a long window is the concern, not a single deliberate skill-broadening move.

### Why do recruiters disagree so much about job hopping?

Because the surveys genuinely split. A LinkedIn survey of 1,024 hiring managers found 37 percent might pass on frequent job changers, a survey of 200 recruiters found 52 percent call it a red flag, and Resume Genius reported 50 percent of employers do not flag it at all. Without shared thresholds and carve-outs, two reviewers will reach different verdicts on the same history. A written rubric is what closes that gap.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/guides/tenure-pattern-read-2*