The Reapply-or-Move-On Decision, Scored After a Rejection

You have a rejection on record and a live posting in front of you at the same company. This guide is for someone running an active search across many companies who has to decide, fast, whether to spend another application on a company that already said no. It turns that call into five scored dimensions and a single verdict: reapply now, wait a named window, or drop the company - plus the one thing that has to change before any second attempt is worth sending.

Public advice converges on one line: wait three to six months and improve. That line is not wrong, but it is useless for most cases, because it treats an instant auto-reject at the resume screen the same as a final-round loss after four interviews. Those are different decisions with different clocks. Below, each dimension is scored on its own, and the section on how this goes wrong carries as much weight as the method itself.

What the five dimensions are, and why stage sets the clock

The reapply decision is not one judgement, it is five, and the first one changes what all the others mean. The five dimensions are: rejection stage, material change, the reposted req and its timing, warm contact, and time elapsed against known cooldowns. Stage is primary because an auto-reject never reached a human, so cooldown logic does not even apply to it.

Here is the trap in the standard advice. Forums are explicit that a cooldown "only counts if you actually get an interview." An automated or resume-screen reject means your file was likely filtered before a person saw it. Firing again after fixing your keywords is a fresh candidacy, not a repeat performance in front of the same audience. So the honest cooldown for an auto-reject is close to zero, while a final-round loss earns you months.

Read the stack top down when you score. If stage says auto-screen and material change says yes, you can often stop there and reapply. If stage says final-round loss, the lower dimensions decide whether the wait is worth serving at all.

Dimension one: rejection stage, read from timing

Classify the stage before anything else, because it sets your cooldown and tells you whether to fix keywords or fix substance. The cleanest tell is how fast the rejection arrived.

A rejection that lands instantly or within a day or two usually means an automated filter screened you out. One that arrives after three or more days usually means your resume reached a human. This is not perfect - some companies batch rejections - but it is the fastest read you have, and it is correct often enough to act on.

One caution worth holding: in a study of 25 recruiters, 92 percent said their applicant tracking system does not auto-reject on formatting or content. So "the ATS killed me" is often the wrong story. A fast rejection more often means a human screener made a quick keyword judgement, or a knockout question (location, work authorization, years of experience) filtered you. Fixing formatting will not help either case. Fixing the substance the screen was looking for will.

The consulting window looks brutal until you see the payoff: roughly 20 percent of current major-firm consultants landed their offer on a later attempt. The long wait is real, but a second attempt there is a genuinely different candidacy, not a formality.

Dimension two: material change, the pass or fail gate

Material change is a gate, not a slider: without a concrete, nameable change since your last application, the verdict is keep improving, not reapply. This is the single rule every named source agrees on.

The test is a question you answer out loud: what can you now show that you could not before? Three answers qualify.

• A new skill relevant to the role, backed by a project you can point to.
• A new project deployed or documented since the last application.
• A new credential, or a target role that has been reconfigured to fit you better.

If you cannot name one of those, you are not ready. "Six months have passed, so I am eligible" is not a change. Reapplying with the same resume and cover letter will not improve your chances, regardless of how much time went by. The clock and the change are two separate dimensions, and passing one does not pass the other.

There is a hidden bonus here. A reported Meta cooldown was cut from 12 months to 6 after the candidate emailed the recruiter and showed concrete progress. That means material change can shorten time elapsed. The two dimensions are linked in your favor: the artifact that makes you a fresh candidate is also the evidence that can renegotiate the wait.

Dimension three: the reposted req, read in days

A repost is a cheap signal most readers misuse; the gap between your rejection and the repost is the whole read. Measure it in days before you spend anything.

Start by ruling out noise. Listings auto-renew on LinkedIn to give the illusion of updated content, and that means nothing. Confirm the post is a genuine reopening before you treat it as a second chance. Then classify it.

• Under one week since your rejection. The search most likely just continued and you were never a serious contender for this instance. A second attempt is usually not worth it here unless a warm contact opens a side door.
• A meaningful gap, weeks or months. The chosen candidate may have fallen through, or the team ended up dissatisfied with all finalists. The prior decision may be reversible. This is the repost worth acting on.
• Changed description. They may now want a different skill set. The original decision repeats unless you closed exactly that gap - which folds straight back into the material-change dimension.

Watch for the evergreen trap: a months-old visa, PERM, or ghost listing read as live demand. If the same req has cycled for three or more months, treat it as background noise, not an opening.

Dimension four: warm contact, the highest-leverage move

A warm contact is worth more than any waiting strategy, and the numbers are not close. This dimension can override a mediocre score on every other dimension.

The whole timing debate is about nudging cold odds that sit at roughly 2 to 3 percent. A referral moves the interview rate to 40 to 60 percent. That is a 13 to 20 times difference on the one metric that matters after a rejection: reaching a human who can reconsider you.

Referrals are only 6 to 7 percent of all applications but account for 30 to 50 percent of hires, with Jobvite data putting hire rates at roughly 30 percent for referred candidates versus 7 percent for other sources. After an auto-reject specifically, reaching a human is the recommended path when the portal is a dead end: if you have a connection inside, contacting them or the recruiter directly is your best option.

The window debate is about moving cold odds a few points. A single referral moves them by an order of magnitude.

How reachable that human is depends heavily on your market. In Refolk's index of professional profiles, there are about 112,900 reachable recruiter and talent-acquisition profiles in the US against about 7,350 in the UK.

The practical read: a US candidate rejected by a portal has roughly 15 times denser a bench of humans to approach directly. UK candidates should weight the warm-contact dimension even more heavily, because the pool to route around an ATS is thinner and each contact is worth more.

Finding the right person is where Refolk removes the friction this paragraph just described: instead of scrolling a company page hoping someone is reachable, you describe the person and get a shortlist back.

Dimension five: time elapsed against known cooldowns

Time elapsed produces a named eligibility date, but the underlying numbers are reported, not published, so treat the date as a working estimate you confirm with the recruiter. No official employer HR page publishes these windows.

The most-cited breakdown from practitioner forums: Amazon around 6 months, and Google, Meta, Apple and Microsoft around 12 months. Others in the same discussions say 6 to 12 months as a range. Consulting runs 12 to 24 months. For anything else, default to 3 to 6 months after any interview stage. The clock usually starts on your rejection date, not on your last contact.

Because material change can shorten this window, do not serve the full wait passively. If you hit a real artifact partway through, email the recruiter, show the progress, and ask whether the window can move. One candidate reported cutting a 12-month wait in half exactly this way.

Run the scoring: the eight-step procedure

Score one rejection in under two hours of active work, spread across a few days while you wait on feedback. Follow the steps in order, but note that some sources treat securing feedback as the very first move and others treat stage classification as primary; do both early, in whichever order fits your reply times.

Reading the five scores into one verdict

Use a simple gate order. First, material change: if it fails, the verdict is drop or wait until you have the artifact, no matter what else scores well. Second, warm contact: if you have a reachable human, reapply now regardless of cooldown, because a referral routes around the portal entirely. Third, stage and cooldown together: if stage is auto-screen and keywords are fixed, reapply now; otherwise serve the named window. Fourth, opportunity cost: if you have already spent heavily on this company, raise the bar for a yes.

Company / role:
Rejection stage: [auto-screen / single round / final-round loss]
Material change (named artifact): [artifact, or NONE - if none, stop]
Repost timing: [auto-renew / continued search <1wk / changed-spec / no repost]
Warm contact: [named person + channel, or NONE]
Eligibility date (cooldown): [date, or N/A for auto-screen]
Applications already spent here: [n]
VERDICT: [reapply now / wait until <date> / drop]
The one thing that must change first: [single sentence]
Trigger to revisit: [event or date]

Dimension six: opportunity cost, the reason to drop

Persisting on one rejecter is measurably costly past a point, so opportunity cost is the dimension that produces a clean drop verdict. Every repeat attempt at one company is an application you did not send elsewhere.

The baselines: most candidates need roughly 20 to 80 applications per offer, 30 to 60 in tech, at a sustainable pace of 10 to 20 quality applications per week. Most searches produce an offer within two to four months, though median time to first offer rose 22 percent to 68.5 days in one recent quarter, so budget generously. Experts say reassess your whole strategy after 30 to 50 applications with no results.

More volume is not free. Beyond 81 applications, offer probability drops to about 20 percent, down from about 31 percent in the 21-to-80 range. So sending more than 80, or firing repeatedly at one company, actively decreases your chances compared to a moderate, targeted volume.

The clean rule: if a company has already cost you multiple attempts and you cannot name a material change or a warm contact, the opportunity cost dimension alone justifies dropping it. There is no published figure for the exact drag of fixating on one rejecter - that is funnel math, not a measured number - but the direction is not in doubt.

How this goes wrong: failure modes and false positives

Most bad reapply decisions come from misreading one signal, and the same handful recur. Each has a false positive and a specific check.

• Misreading a repost as a second chance. An auto-renewed LinkedIn listing read as a reopened req. Check: confirm the gap since your rejection and compare the descriptions before acting. Auto-renewal means nothing.
• Reapplying identically. "Six months passed, so I am eligible" - with the same file. Check: can you name the artifact that changed? If not, the wait did nothing.
• Assuming an ATS killed you when a human did. A three-or-more-day rejection means content, not keywords. Fixing formatting will not help; fix substance.
• Treating final-round feedback as a re-openable negotiation. After detailed final-round feedback, the decision was usually deliberate. A quick reapply without change is unlikely to move it.
• Taking cooldown numbers as policy. The FAANG windows are forum reports. Verify with the recruiter, since the clock usually starts on the rejection date.
• Over-indexing on volume to beat the rejecter. Past about 80 applications, odds fall. More shots at one company is worse, not better.
• Chasing a repost that is a visa, PERM, or ghost listing. A months-old evergreen post read as live demand. Check whether the same req has cycled for three or more months.

The most expensive of these is the identical resubmission, because it burns the material-change gate and the warm-contact opportunity at once: you spend your one reintroduction to a company on a file they already declined. The check is trivial and the cost of skipping it is your candidacy.

Keep the decision current and close it out

Before you act on any verdict, run this check so the decision holds up when a recruiter reopens your file.

Keep the decision alive rather than filed and forgotten. Cooldowns are estimates, so recheck the eligibility date against anything the recruiter tells you, and reset the clock only from a confirmed rejection date. Reposts change, so re-run the day-gap read if the listing reappears. And revisit the material-change line as your search continues: the project you ship next month may flip a wait verdict into a reapply, and may be the exact evidence that shortens the window when you email the recruiter to say so.

When you reach the warm-contact step and cannot see a route in, that is the moment to widen the search for a human rather than resubmit through the portal that already rejected you. Refolk lets you describe the recruiter or the former colleague you need and returns reachable profiles, so the highest-leverage dimension does not stall on not knowing who to ask.