The Skip, Quick, or Full Posting Triage Rubric

You have a posting open in front of you. The job is to decide, in the next five minutes, whether to skip it, fire off a quick application, or invest an hour in a tailored resume and cover letter. This guide is the single scoring rubric that makes that call, for job seekers running an active search across many companies at once. It combines the two halves the internet keeps separate - "always tailor" advice and "here are the ghost-job red flags" - into one per-posting decision that also tells you how much effort to spend.

The reason this matters is arithmetic, not taste. Tailoring effort should follow fit and liveness, not be spread flat across every application or wasted on dead listings. Below is the rubric, the four dimensions it scores, the tiers it routes into, and the specific ways it lies to you.

Why one rubric beats "tailor everything" or "spot the ghost"

The two dominant answers are each half right and together incomplete. "Tailor every application" ignores that a large share of postings will never hire anyone, so the effort lands on dead air. "Here are six ghost-job red flags" tells you what to avoid but not what to do with the live roles that remain, or how hard to work each one.

The gap between them is a routing decision. You are not answering "is this a ghost?" in isolation and you are not answering "should I tailor?" in isolation. You are answering both at once, for the posting in front of you, and reading the combined answer into an effort level.

The market data justifies the split. Full tailoring costs roughly two to four times the time of a keyword-only pass, and it roughly doubles the interview rate, from 2.68% to 5.75% in a 59,000-resume analysis. But that lift only exists if a human reads the application. Spend 60 minutes on a role that has a 27% chance of being a ghost and the expected return collapses.

Full effort only pays where liveness is confirmed, so score the listing before you score yourself.

The prevalence estimates disagree wildly, and that disagreement is itself a triage argument. They measure different things: employer-admission surveys, ATS internals, and gaps in national hiring data. A per-posting rubric sidesteps the whole debate by scoring the listing in your hand rather than betting on a market-wide rate.

How bad is the ghost-job problem, really

Bad enough that ignoring liveness wastes a large fraction of your effort, but the exact rate depends entirely on how you measure it. The honest answer is a range, and the range is the reason to score each posting individually.

Read the table carefully before you panic. The 93% figure is not "93% of jobs are fake" - it is the share of HR professionals who admit their employer posts ghost jobs at all, with 45% saying regularly and 48% occasionally. The listing-level rate is closer to one in four or one in three. In June 2025, US employers reported 7.4 million openings but made only 5.2 million hires, a gap of more than 2.2 million.

Sector changes the picture. Government postings run near a 60% ghost rate, education and health near 50%, and information and tech near 48%. A 45-day-old tech pipeline post is more suspicious than a 45-day construction role, where hires still roughly match openings. Fold the sector into your age read rather than applying one threshold everywhere.

The four dimensions you score

Every posting gets scored on four things, in this order: liveness, fit, timing, and access. Liveness comes first because it gates everything - a dead role at 100% fit is still worth zero.

Liveness is whether a human will actually read your application. The signals are posting age, repost cadence, first-party corroboration, salary-band presence, and recent employer activity. Genuine roles typically fill within 30 to 60 days, so a listing over 60 days old with no description changes is high-ghost risk. If the same role reappears every few weeks with a fresh date, the position has not been filled and may never be.

Fit is your coverage of the posting's required keywords. Separate the must-haves from the nice-to-haves, then estimate what fraction of the required terms you actually hit. A posting typically holds 15 to 25 key terms worth matching.

Timing is freshness and momentum. A three-day-old post on a growing team beats a fresh post at a company that just announced layoffs.

Access is whether you have a referral, a named contact, or any way past the resume pile. Access is the single largest multiplier on a live, well-fitting role, and it is the reason some Quick candidates become Full.

Reading required vs preferred without over-penalizing yourself

Required qualifications are must-haves, and missing them significantly hurts your score. Preferred qualifications are bonus points that improve your ranking but are not gates. Over-weighting the nice-to-haves inflates your perceived misfit and causes needless skips. The "preferred" or "nice to have" terms are exactly that.

Coverage floors vary by source. One recommends 90% of required skills present and 70% of preferred; another says cover 75 to 80% of required keywords and all core qualifications; a looser apply-threshold view says apply if you match 60 to 70% of required. Pick one floor and hold it across your search so your tiers stay comparable. Mirror the exact required strings, because an ATS may not recognize "client satisfaction" when the posting says "customer success."

What each tier costs and what it returns

Three tiers, three effort levels, and the math behind why they exist. Skip costs nothing. Quick is a keyword-only resume pass. Full is a tailored resume plus a cover letter.

The keyword-only pass runs 3 to 5 minutes with tooling or 20 to 40 minutes by hand, and involves roughly 5 to 10 line-level changes to the resume. Full tailoring runs 30 to 90 minutes, with diminishing returns beyond that. Recruiters cap the sane cover-letter spend at 20 to 30 minutes, so do not let the letter swallow the whole budget.

The derived column is the point. Full effort roughly halves your applications-per-interview, from about 37 to about 17. That is a real return, but it is a return you only collect when the role is live. This is where a tool earns its place: Refolk writes your resume from your own history and tailors it to a specific posting, which collapses the 20-to-40-minute manual keyword pass into the tooled range and frees the saved time for the liveness vetting that actually protects the return.

The five-minute triage procedure

Run these steps in order on the posting in front of you. The whole pass is designed to fit inside five minutes, and it stops early if liveness fails.

The metadata step is worth defending. Posting age and repost history carry about 26% of one open ghost-job detector's weighted score, yet no major board fully discloses repost history. The single most valuable signal is the one you have to dig for, which is exactly why a manual liveness step beats trusting the board's UI.

How this rubric lies to you

The rubric is only as good as your skepticism about its signals. Each dimension has a failure mode where the signal points the wrong way, and skipping this section is how careful people still waste weeks.

• Age read as death. A 60-day-old post can be a real role stalled by a spending freeze. A role approved in Q1 budgeting gets posted, then a freeze hits in Q2, and the listing stays up because nobody takes it down. False positive: skipping a live role. Check whether it is still on the careers page and whether a recruiter touched LinkedIn recently.
• Careers-page presence read as liveness. Boards and aggregators lag, so the role may already be closed at source. Check the original date on the first-party page, not on the board that surfaced it.
• Keyword coverage read as fit. Hitting 80% of terms through synonyms fails ATS exact-match. If the posting says "customer success" and you write "client satisfaction," the parser may miss it. Mirror the exact required strings.
• Preferred treated as required. Over-weighting nice-to-haves inflates perceived misfit and triggers needless skips. Score required and preferred separately and let only the required floor gate you.
• Fast auto-rejection read as your weakness. A rejection that arrives suspiciously fast often means the application vanished into nothing, not that it was weak. Check your own log for a pattern per company before you rewrite your resume.
• Missing salary band treated as neutral. It is a liveness and quality signal, not a neutral one. At scale, 64% of Canadian workers skip postings with no salary. Count a missing band against liveness.
• "No public hiring signal" read as not-hiring. In Refolk's index, only 0.02% of US recruiters publicly say "actively hiring," so the absence of that headline is not diagnostic. Read recruiter activity and employee-count growth instead of headline text.

There is a volume failure mode too, on the other side. A realistic plan includes 10 to 25 targeted applications for serious roles, not 3 to 5. Reserving Full effort for too few live roles under-samples the market. The rubric is meant to raise your Full-effort hit rate, not to shrink your search to a handful of applications.

Using recruiter supply as a liveness backdrop

Individual postings do not disclose whether anyone is hiring, but the recruiter population around a role does. Refolk's index gives a backdrop for reading that.

The takeaway is not the raw counts, it is the 23. Public hiring declarations are so rare that their absence carries no information. The usable move is to look for recruiters and hiring-team members who have been active recently, and to check whether the team the role reports into actually exists and is growing. That is a people question, and it is where a search tool that finds and confirms real humans beats staring at a job board.

Finding one engineer already on the team the role reports into, or a recruiter who left the company in the last six months, tells you more about liveness than the posting's own copy ever will. A recent departure can flag a backfill rather than a ghost; a growing team confirms the role is real.

Before you call the triage done

Run this checklist against the posting before you commit to a tier. It is the difference between a rubric you follow and a rubric you feel like you followed.

Keeping the rubric current

Your thresholds are hypotheses, and your log is how you test them. Track date, posting age, tier, and response for every application, because that record is the only way to see which boards and companies never respond, and to recalibrate.

Re-check your floors against your own outcomes rather than against any single source. If your Full-effort applications convert well above the 5.75% benchmark, your liveness vetting is working and you can push more roles into Full. If Quick applications to a particular board return nothing but fast auto-rejections, that board is failing your liveness screen and belongs lower in your rotation.

The market-wide numbers here will drift, and the sensible response is to re-check the mechanism, not memorize the value. Ghost-job prevalence, sector rates, and interview yields all move; what stays fixed is the shape of the decision. Score liveness first, let it gate fit, let fit and access set the tier, and spend the 30-to-90-minute Full effort only where a human is confirmed to be waiting on the other side.