# Sourcing Return-to-Work Candidates From Public Signals

*You can find return-to-work candidates from public signals, qualify them on skills without penalizing the gap, and deliver a ranked, reach-ready shortlist for one role.*

- Canonical URL: https://www.refolk.ai/guides/sourcing-return-to-work-candidates
- Pillar: Recruiting and sourcing
- Format: Playbook
- Published: 2026-10-06
- Last reviewed: 2026-10-06
- Reading time: 16 min
- Keywords: how to source career break candidates, sourcing return to work professionals, finding candidates with resume gaps, returnship candidate sourcing, return to work talent pipeline

## Key takeaways

- In Refolk's index, 1,340 UK professionals carry a Career Break title versus 617 in the US - 2.17x more in a smaller workforce, so a UK-first run yields a denser pool per hour.
- Filtering Career Break profiles by seniority returns near-zero because the break entry overwrites the title field; in Refolk's index, 0 of 617 US break-labelled profiles showed at Director or above.
- Warm re-engagement runs 25-40% positive response versus 10-20% cold, and re-engaging past candidates speeds hiring by 50-70%, so split warm from cold before you write a single message.
- Gap-length knockouts are a legal liability, not just a missed candidate: they disproportionately hit over-40 and caregiver applicants, and the EEOC's Mobley v. Workday position extends liability to the screening tool itself.
- Use the published eligibility floor - roughly 5 years prior experience and a 6-month-to-3-year gap - as inclusion criteria, never as knockouts.
- A four-touch sequence generates 2x the replies of a single email, and performance flattens after the fourth stage, so plan exactly four touches over 10 to 14 days.

This is a procedure for building a ranked shortlist of experienced professionals returning from a career break, tied to one open requisition, found entirely from public history signals. It is for in-house recruiters, sourcers, and founders hiring directly who want a repeatable sourcing run rather than a diversity-hiring intention. By the end you can surface returners from public signals, qualify them on demonstrated skill without penalizing the gap, and hand off a reach-ready list ordered by evidence.

Most published content on career breaks either helps job seekers find returnship programs or helps employers design one. This does neither. It is the sourcing side: how to go find and rank individual returners on a Tuesday morning, for a req that is open now.

## What counts as a return-to-work signal, and where each one lies

A return-to-work signal is any public marker that a person stepped away from paid work and is now re-entering. The strongest explicit one is LinkedIn's Career Break feature, launched March 1, 2022, which lets a member log a break in the Experience section by selecting from 13 reasons such as caregiving, bereavement, travel, or retirement. These entries are searchable inside LinkedIn Recruiter, which is what makes this feature a sourcing tool rather than a profile cosmetic.

The secondary signals are weaker but wider in coverage: a dated experience gap with no entry in it, and a headline that says "returning to work" or "relaunching." Each signal proves something slightly different, and each one lies in a specific way.

| Signal | What it proves | How it lies |
|---|---|---|
| Career Break entry | The person self-identified a break and is open about it | Overwrites the title field, so seniority reads as blank |
| Dated experience gap | A pause in logged paid employment | May have been freelance or consulting the whole time |
| "Returning to work" headline | Active intent to re-enter now | Often set long before the person is truly available |

The lesson in that table is that no single signal is sufficient. The Career Break entry is the cleanest intent marker but strips out seniority. The dated gap catches people who never used the feature but cannot distinguish a true break from an unlogged consulting stretch. Path Forward is explicit that freelance or consulting work during a break generally should not count against a candidate, so a bare gap is an invitation to read the narrative, not a verdict.

One more caution on the explicit label. A practitioner analysis reported, unverified, that using the Career Break feature may suppress a profile in some recruiter searches. I cannot confirm that from the dossier, and I would not plan around it. But it is a reason to run the dated-gap and headline searches in parallel rather than leaning on the labelled pool alone.

## Where the returners actually are: UK versus US

Returners are not evenly distributed, and the gap is large enough to change where you start. In Refolk's index, 1,340 UK professionals carry a "Career Break" title against 617 in the US - 2.17x more in a smaller workforce. That points to higher feature adoption and openness in the UK, which means a UK-first sourcing run gives you a denser labelled pool per hour of work.

**2.17x - UK versus US professionals with a logged Career Break title**

Refolk's index shows 1,340 in the UK against 617 in the US, a denser labelled pool in a smaller workforce.

| Country | Career Break profiles | Multiple |
|---|---|---|
| United Kingdom | 1,340 | 2.17x |
| United States | 617 | 1.0x |

Do not read these as the total addressable pool. They are only the explicitly labelled slice. The much larger population of returners sits behind dated gaps and summary-text descriptions that the label search never touches. Treat the labelled count as your fastest seam to mine first, then widen to the implicit signals once the obvious pool is worked.

If your role is UK-eligible or remote, start there. If it is US-only, expect to lean harder on dated gaps and headlines, because the clean labelled pool is smaller and, as the next section shows, it hands you almost no seniority data.

## The eligibility floor: use it to include, never to exclude

Published return-to-work programs converge on a narrow eligibility band, and you should borrow it as an inclusion filter. The common floor is roughly 5 years of prior professional experience and a gap somewhere between 6 months and 3 years. Here is how named programs set the two dials.

| Program | Duration | Min prior experience | Min gap |
|---|---|---|---|
| Expedia Return Ticket | 16 wks | 5 yrs | 2 yrs |
| VMware Ascent | 18 wks | ~5 yrs | ~2 yrs |
| Intel | 16 wks | 5 yrs | 1 yr |
| Dell | n/a | 5 yrs | 12 mo |
| GM Take 2 | 12 wks | n/a | 2 yrs |
| Bank of England | 12 mo | required | 2 yrs |

Read the columns and the pattern is clear: experience floors sit at about 5 years and gap floors run from 12 months to 2 years, with the Bank of England Career Returner Programme requiring a minimum two-year break before the start date. iRelaunch, which has maintained a return-to-work directory since 2011, defines relaunchers as mid-to-senior professionals who took a break of one to two years or more.

The critical move is directional. Use these numbers to decide who qualifies as a returner worth sourcing, not to knock anyone out. A gap-length knockout does two bad things at once: it discards high-skill long-break candidates, and it creates disparate-impact exposure I will come back to. SHRM reports that applicants with work gaps had a 45% lower chance of interviews than those without, and Path Forward notes interview chances drop by roughly half once a gap exceeds two years. Those numbers describe the bias you are routing around, not a rule to replicate.

> **Rule:** Eligibility is a floor, not a ceiling
>
> Apply the 5-years-and-6-months-to-3-years band only as inclusion criteria. The moment you use gap length to reject someone, you have rebuilt the exact bias that makes returners hard to hire.

## Qualifying on evidence, not on recency

Qualify a returner on demonstrated competence, not on how recently they were paid. The documented skills-based method starts from what the person must do in the first 90 days, breaks the role into tasks, identifies the skills each task requires, and then measures those skills with work samples, structured interviews with standardized rubrics, technical assessments, and portfolio reviews.

This is not a soft preference. Skills tests carry far higher predictive validity than the proxies gap screening relies on: general mental ability assessments score around 0.65, against 0.07 for years of experience and 0.10 for education. In other words, the thing a gap filter protects - continuous tenure - is one of the weakest predictors of performance you could pick.

> A candidate's last payroll date should not outweigh ten years of proven ability.

For returners specifically, give greater weight to total relevant experience, verified competence, training, and any upskilling done during the break. The returnship model itself is a work-sample evaluation in disguise: programs let the employer judge a relauncher on observed work instead of an interview, deferring the permanent decision until the end. That is why completers convert at over 80% at Path Forward and 87% at reacHIRE, with a 92% two-year retention rate reported by reacHIRE. The mechanism is removal of interview recency bias, and a direct-hire sourcer can borrow it with a paid work sample instead of a tenure check.

Treat those conversion figures with care, though. They are program-completer rates, not sourced-pool rates, so they tell you the method works for people who finish, not what share of your raw pool will finish.

#### From role to a qualified returner

1. **Outcomes** - Define first-90-day outcomes for the role
2. **Tasks** - Break each outcome into tasks
3. **Skills** - Name the skill each task needs
4. **Evidence** - Score candidates on work samples and portfolios, not recency

*Each stage converts a vague intention into a specific, evidence-backed judgement.*

When you build the search, this is where friction appears: the labelled pool strips seniority, the implicit pool hides inside free text, and you need to read narratives at volume. [Refolk](/) lets you ask for the pre-break role and the skill in one request, so you recover the seniority the Career Break label erased.

I ran this search: `Find people in the UK with a logged career break who have 5+ years in product management and shipped a B2B SaaS product before their break.` - [see the full result list](https://www.refolk.ai/s/meb55dfcqt).

*Returns returners with the pre-break seniority and a concrete shipped-product data point, the two things a raw label search drops.*

## The step-by-step sourcing run

Run these seven steps in order for a single requisition. The time estimates assume one recruiter and one sourcer; the elapsed clock is dominated by the outreach sequence, not the desk work.

#### One requisition, start to shortlist

1. **Scope the requisition to a skills spec** - With the hiring manager, translate the role into first-90-day outcomes and the skills each requires. Done is a ranked must-have skills list and three outcome statements, with no continuous-years line. (~1-2 hours)
2. **Build the signal search** - Query for the Career Break entry plus dated gaps and "returning to work" headlines, filtered by the skills spec. Done is a raw pool tagged by signal type. (~1-2 hours)
3. **Apply the eligibility floor, not a ceiling** - Use roughly 5+ years prior experience and a 6-month-to-3-year gap as inclusion criteria only. Done is a pool split into "fits returner profile" and "needs review," with no long-gap knockouts. (~1 hour)
4. **Qualify on evidence** - Score each candidate on work samples, portfolios, pre-break accomplishments, and break-period upskilling, not recency. Done is every shortlisted profile carrying at least two skill-evidence data points. (~2-4 hours)
5. **Split warm from cold** - Flag silver-medalists, alumni, and prior applicants as warm; route the rest to cold. Done is two lists with response benchmarks attached: warm 25-40%, cold 10-20%. (~30 min)
6. **Run the re-engagement sequence** - Four touches across LinkedIn, email, and SMS over 10-14 days, leading with a soft 15-minute-call ask. Done is the sequence complete and replies logged. (10-14 days elapsed)
7. **Rank and hand off** - Order by skill-evidence score multiplied by engagement signal. Done is a reach-ready ranked shortlist with an evidence note per candidate. (~1 hour)

There is one defensible reordering. iRelaunch frames the work-sample evaluation - the equivalent of step four - as the hiring decision itself rather than a pre-screen. Direct-hire employers who are not running a formal returnship tend to compress steps four and seven, scoring evidence at the same moment they rank. Either way the logic holds: evidence before recency, inclusion before exclusion.

## Warm versus cold, and the four-touch sequence

Split your pool into warm and cold before you write anything, because the two convert at roughly double the rate of each other. Cold but targeted outreach lands a 10-20% positive response; warm introductions and community contacts land 25-40%. The midpoints put warm at about 2.2x cold.

| Channel type | Positive response | Source |
|---|---|---|
| Cold, targeted | 10-20% | sapia.ai |
| Warm / community | 25-40% | sapia.ai |
| 1st touch (Gem) | 8.3% | pin.com |
| 5-touch cumulative (Gem) | 21.3% | pin.com |

Your warm pool is silver-medalists, alumni, and prior applicants. Re-engaging past candidates can speed hiring by 50-70%, and returners reward a prior relationship disproportionately because, as iRelaunch notes, relaunchers apply only after significant thoughtfulness. A deliberate candidate responds to someone who already knows them.

For the message itself, the data is unambiguous about depth. A four-email sequence generates 2x the replies and a 68% higher interested rate than a single email, and performance flattens after the fourth stage, so there is no reason to plan five. Gem's figures show the compounding directly: a single touch replies at 8.3%, while a five-touch sequence reaches 21.3% cumulative. Lead with a soft ask - a 15-minute exploratory call rather than a formal interview - which consistently lifts response because it lets passive talent engage without feeling locked into a process.

```funnel
title: Reply compounding across touches
caption: Most replies arrive after the first message, so a single touch disqualifies no one.
stage: First touch :: 8.3% :: single-email reply rate (Gem)
stage: Full sequence :: 21.3% :: five-touch cumulative reply rate (Gem)
```

**Day 1 LinkedIn connection note for a returner (under 400 characters)**

```
Hi Priya, I came across your work leading product at a B2B SaaS team and the launch you shipped there. I'm hiring for a senior product role and your background is a strong match. Open to a quick 15-minute call to compare notes, no process pressure either way?
```

*Reference the pre-break role and the skill, never the gap or its reason. Swap the bracket-free specifics for your own role before sending.*

A workable cadence: Day 1, the LinkedIn connection note above, under 400 characters. Day 3 to 4, an email with role and team context. Day 7, a LinkedIn follow-up referencing the email. Day 10, an SMS or a brief call. Four touches, two weeks, done.

## How this goes wrong

The failure modes below are where returner sourcing runs quietly produce a bad shortlist that looks fine. Each has a false result it generates and a check that catches it.

- **Using the Career Break label as your only filter.** This misses returners who described the break in their summary or logged it as freelance. The false positive it admits is a self-employed consultant who never actually left. Check by reading the gap-period narrative, not just the label.
- **Filtering break-labelled profiles by seniority.** This returns near-zero, because the break entry overwrites the title field. In Refolk's index, 0 of 617 US break-labelled profiles showed at Director level or above. The false negative reads as "no senior returners exist." Check by inferring seniority from the pre-break role.
- **Applying a gap-length knockout.** This excludes high-skill long-break candidates and creates ADEA and disparate-impact exposure, while admitting a "clean" shorter-gap but weaker candidate as a false positive. Check by using the gap only to include, then qualifying on work samples.
- **Asking "why the gap" in outreach.** This risks eliciting protected-class information and a discrimination claim. Check by asking only about skills and outcomes.
- **Sending a single touch to voluntary leavers.** This leaves 40-60% of replies on the table and reads as "they're not interested." Check by completing the full four-touch sequence before disqualifying anyone.
- **Treating freelance income as "still employed."** This wrongly disqualifies a genuine returner, since Path Forward is explicit that freelance generally does not count against them. Check by confirming whether the work was in the target field at the target level.
- **Trusting a conversion stat without its denominator.** The 87% and 80% figures are completer rates, not applicant-pool rates. Check by asking what share of sourced candidates actually reach completion.

**0 - Director+ returners found by filtering break-labelled profiles on seniority**

In Refolk's index, filtering 617 US Career Break profiles to Director and above returned zero, because the label overwrites the title field.

> **Watch out:** The seniority filter will lie to you
>
> If you filter the labelled returner pool by seniority, you will conclude no senior returners exist. They do - their titles are just hidden behind the break entry. Source seniority from the pre-break role every time.

The legal exposure deserves its own line, because it is the failure mode most likely to turn into a filing rather than just a weak hire. No US law names employment gap as a protected basis, and the EEOC says employers cannot discriminate on unemployed status but may ask about it, provided any rejection on that basis is applied consistently. The ADEA protects workers 40 and older, and resume-screening tools that filter graduation years or gaps may disproportionately exclude them. In an April 9, 2024 amicus brief in Mobley v. Workday, the EEOC argued that an AI resume-screening vendor can itself be liable under Title VII, the ADEA, and the ADA. The mechanism that makes a gap filter feel efficient is exactly what makes it a disparate-impact risk.

> **Tip:** Keep the question on the work
>
> SHRM warns that asking about a gap can draw out criminal history, health, family, veteran status, disability, or age - all of which raise discrimination risk. You never need the reason. Ask what the person can do and what they shipped.

## Before you call the shortlist done

Run this check before handing the list to the hiring manager. It catches the failure modes above while they are still cheap to fix.

#### Shortlist release check

- [ ] Every shortlisted candidate carries at least two skill-evidence data points
- [ ] No candidate was excluded on gap length alone
- [ ] Seniority was read from the pre-break role, not the current Career Break entry
- [ ] The pool was drawn from labelled, dated-gap, and headline signals, not the label alone
- [ ] Freelance or consulting during the break was confirmed as in-field and at-level before counting it
- [ ] Candidates are tagged warm or cold with the matching response benchmark attached
- [ ] Outreach copy references skills and outcomes only, never the reason for the gap
- [ ] The full four-touch sequence ran before anyone was marked not interested
- [ ] Any borrowed conversion stat is labelled as a completer rate, not a pool rate

## Keeping the pipeline current

A return-to-work pipeline decays fast because the "returning now" window is short, so re-run the signal search on a cadence rather than once. The explicit Career Break pool turns over as people re-enter employment and their entries age out, and the headline pool turns over faster still, since "returning to work" is often set weeks before someone is truly available.

Three things to re-check and how. First, re-run the labelled search monthly against live reqs, because the UK-versus-US density in Refolk's index tells you which geography repays the time. Second, revisit the eligibility band only if a named program publishes a different floor; the 5-years-and-6-months-to-3-years shape has been stable across Expedia, VMware, Intel, Dell, GM, and the Bank of England, so treat a sudden outlier as a reason to read the source, not to move your filter. Third, keep watching the legal surface: the Mobley v. Workday position is active, and anything that pushes automated-screening liability further changes how you may filter, not just how you should.

The returner pool is larger and more deliberate than the labelled slice suggests, and most recruiters never work the implicit signals at all. The team that reads narratives instead of filtering on gaps, and that runs the full four touches instead of one, converts a pool its competitors have already dismissed.

## Frequently asked questions

### How do I find candidates with resume gaps without a paid tool?

Start with LinkedIn's Career Break feature, which lets members log a break in the Experience section and is searchable inside LinkedIn Recruiter. Layer in dated experience gaps and 'returning to work' headlines, then filter by your skills spec. The label alone misses people who described a break in their summary or as freelance, so read the gap-period narrative rather than trusting the label as your only filter.

### What is a reasonable eligibility threshold for return-to-work sourcing?

Published returnship programs cluster around 5 years of prior experience and a gap between 6 months and 3 years. Expedia and VMware require about 5 years and a 2-year break; Intel and Dell accept a 1-year break. Use these as inclusion criteria to define who fits the returner profile, never as knockouts, because a longer gap does not mean a weaker candidate.

### Is it legal to screen candidates on employment gaps?

No US law names employment gap as a protected basis, but gap screening carries real disparate-impact risk. The ADEA protects workers 40 and older, and gap filters disproportionately exclude older and caregiver applicants. The EEOC argued in its April 9, 2024 amicus brief in Mobley v. Workday that an AI screening vendor can be liable under Title VII, the ADEA, and the ADA. Qualify on work samples instead of gap length.

### Why can't I filter return-to-work candidates by seniority?

Because LinkedIn's Career Break entry sits where a job title would and overwrites it. In Refolk's index, filtering 617 US break-labelled profiles to Director level and above returned zero results. A seniority filter will make you conclude no senior returners exist. Infer seniority from the pre-break role instead of the current entry.

### How many touches should a return-to-work outreach sequence have?

Four. A 2024 analysis of 4 million recruiting emails found four-email sequences generate 2x the replies and a 68% higher interested rate than a single email, and performance flattens after the fourth stage. Spread the four touches across LinkedIn, email, and SMS over 10 to 14 days, and lead with a 15-minute exploratory call rather than a formal interview.

### Do the high returnship conversion rates apply to my sourced pool?

Not directly. The 80% and 87% conversion figures are program-completer rates, measured on people who already entered and finished a structured returnship, not on a raw sourced pool. Treat them as evidence that deferring the decision until after observed work removes recency bias, but ask any program what share of sourced candidates actually reach completion before you borrow the number.

---

*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/sourcing-return-to-work-candidates*