Refolk / Guides
Reference documents for running your own job search: how to write the materials, apply at volume, interview, and negotiate, in a form you can work from rather than skim.
Skip ahead and let me write itDeciding what you are selling and getting it onto a page: the resume, the profile, the portfolio, the story.
Turn your last salary and benefits into a target and floor for W-2, 1099, and C2C hourly work, plus one rate line to send a recruiter before submission.
Dividing a $120,000 salary by 2,080 hours gives $57.69 an hour, which earns about 7.3% less than the salary once you bill only 1,928 hours after 8 holidays and 11 vacation days.
Grade each attainment, quota, rank, club, and earnings line Pass or Fail on seven written criteria, so it survives a W-2 back-solve without you sharing pay history.
On a 50/50 plan with $200K OTE, 118% attainment raises total pay only about 9%, to roughly $218K, so a manager who compares attainment to total pay instead of variable pay will see a gap that isn't real.
Score one short-tenure job on six weighted dimensions, land on List, Compress, or Omit, and leave with a matching LinkedIn entry and a written rule for every disclosure form.
Score a short stint against the BLS median for your own age band and occupation: three months is about 17% of the 1.5-year median for ages 20-24 but only about 3% of the 9.6-year median for ages 55-64.
Turn one employer's filings, call transcript, posting and team blog into a timed 60 to 90 second answer, with a recruiter cut and a hiring-manager cut you can audit.
A strong why-this-company answer has two traces: every claim about the employer points to a dated public source, and every claim about you points to a line on your resume.
Hear any self-pitch prompt, find its row, name what it grades, and answer from the right resume line at the right length for the stage you are in.
Self-pitch prompts fall into five rows (open pitch, resume walkthrough, differentiator, strongest case, not on your resume), and each one grades something different: fit, career-move logic, difference, case, or character.
Score every written field in one application on five dimensions, give each a Send, Rework, or Hand-Write call, and add the disclosure line wherever the rule requires one.
Published employer AI rules mostly police whose content an application is, not whether a tool touched it: the UK Civil Service, DWP, Anthropic and UKRI all allow refining and ban generation.
Look up any scope figure in your history and name its leveling criterion, how it misreads at small and large employers, and the wording that places it in your target band.
In Refolk's index there are 1.9 US Engineering Managers for every US Director of Engineering (43,215 to 22,741), yet Pave finds directors average fewer direct reports (4.6) than managers (4.9), so the roll-up org carries a director claim and the direct-report count does not.
Reach a written Senior or Staff verdict backed by dated evidence on five axes, mapped to each employer's level names and confirmed on the first recruiter call.
In Refolk's index of professional profiles there are about 5.5 US Senior Software Engineers for every US Staff Software Engineer, so a recruiter's default assumption is Senior unless your materials prove otherwise.
For each job profile you keep, score the pay field on four dimensions and record one verdict (target, floor, or blank) with the exact figure and a review date.
On Indeed, the profile minimum pay is hidden from employers even when the profile is searchable, and it only highlights matching jobs, so entering your full target there carries no pricing risk.
Grade your portfolio review deck pass or fail on seven gates, with time thresholds for 30-minute and 45 to 60-minute slots, before you share your screen.
Time a portfolio deck against the talking time the recruiter states, not the calendar slot: a 55-minute call with ten minutes reserved for questions gives you 45 minutes to present.
Build one page naming your target, three title families, 15 to 40 companies, and three proof points, then derive three versions a stranger reads correctly.
In Refolk's index, "Product Owner" has 14,505 US profiles against 67,546 for "Product Manager", a ratio of 0.21, and only 64% of a 25-profile Product Owner sample carried that exact title. An alternate title is a separate market with its own company list.
Take one public-sector posting and answer every self-rating and written prompt so each scores on its own, survives the audit, and targets a reachable rank.
California training and experience evaluations are usually a scored component worth 100% of the rating, so the questionnaire sets your score, not your resume.
Look up any discoverability toggle on LinkedIn, Indeed, Monster, SEEK, Dice, Wellfound, Naukri or cord, know who it really exposes you to, and audit every account in about an hour.
In Refolk's index, 29,254 of 122,941 US recruiters (23.8%) work at staffing and recruiting firms, and employer-exclusion filters keyed on company identity do not cover them.
In one three-hour session, build answers for 8 to 10 recurring prompt families at three lengths, each traced to a record, then fill any form in under ten minutes.
Greenhouse single-line answer boxes stop at 255 characters and cut longer text mid-sentence with no error, and the public job-board API does not show that limit, so a pre-written 255-character version is the only safe default.
Grade every experience bullet pass or fail on seven fixed tests and every role block on lead and count, with verdicts a second reviewer will reproduce.
An experience bullet passes only if it clears all seven binary tests: title-inferable, owned action, result or substitute evidence, baseline, specific noun, length, and sourced number.
Grade any cover letter on seven pass/fail criteria, so two readers reach the same send-or-fix verdict and the same list of fixes in about fifteen minutes.
After an AI drafting tool became available on a large freelance marketplace, the correlation between a cover letter's alignment with the posting and its callback rate fell by 51%, so matching the posting's language no longer sets a letter apart.
Take any resume and one Title 5 announcement to a two-page file with every required field, a line behind each clause and answer, and a logged cut count.
USAJOBS requires five items per job (employer, title, month/year dates, hours per week, and how you meet the qualifications) plus series and grade for federal jobs, and its Help Center list does not include street address, supervisor or salary.
For any optional field on an application form, decide in under a minute whether to fill it, with which artifact, and what a blank or broken entry signals.
In Refolk's index of professional profiles, 889 US recruiters tag Workday as a skill against 236 who tag Greenhouse, a 3.8x gap, so forms with employer-set upload caps are the more common case.
Turn 30+ raw experiences into 8 to 12 sourced, probe-ready stories, with every target competency covered at least twice and every gap named with a fill plan.
A behavioral story bank should hold 8 to 12 stories, because structured interviews typically assess 4 to 6 competencies and two stories per competency leave room for a weak fit or a repeat across a loop.
Grade each LinkedIn and job-board profile on gate, rank, and response tests, reach the verdict a second grader would, and list the exact fields to fix.
In Refolk's index, adding a single SQL skill filter to a US "Data Analyst" title search cuts the pool from 65,307 to 12,914 profiles, so 80.2% of title matches disappear when the skill isn't in the Skills section.
Running a search as a process: where to find openings, how many to send, what to tailor, and how to keep track.
For any application form question, look up whether it can auto-reject you, how to count your facts against it, and what a rejection's timing reveals.
Six applicant systems document an auto-reject rule tied to a screening answer: LinkedIn, Indeed, Greenhouse, iCIMS, Taleo, and Workday, under six different names.
For every live posting, pick a referrer, call refer-first or apply-first, send a packet, and confirm by day 7 that the record shows a referral source.
Most of the referral advantage comes before the interview: in a Federal Reserve study, referred applicants had about 2.2 times the interview probability of job-board applicants, against a 14 percentage point offer gap after interview.
Turn any posted pay range into a new-hire landing zone, place your target in it, and sort the posting to Send, Level-Check, or Skip with a written reason.
A single professional pay band usually spreads 40 to 50% above its minimum, so a posted range wider than about 65% probably covers two adjacent levels, and one wider than about 73% is wider than two standard grades stacked together.
Turn one researched pay range into a fixed entry for all eight salary field types, check each against the posted range and local law, and log it for the screen call.
On Greenhouse, only yes/no, single-select and multi-select questions can trigger auto-reject, so a salary dropdown band carries more filtering risk than a typed number in a free-text box.
Take any posting with an AI-matching notice and say, per resume input, what the matcher credits, how it misreads it, and the one edit that fixes it.
On Workday HiredScore, a B grade means you meet all basic qualifications, and missing even one basic qualification caps you at C no matter how many preferred qualifications you cover.
Split six screens into one shared core and six capped delta sheets, log every minute, and name after the week which delta items the calls used and which were wasted.
At Ashby's 35% recruiter-screen pass-through, six screens produce about 2.1 advances, so roughly four of six per-company prep sheets go to processes that end at the screen.
Grade any resume file pass or fail on six field-level criteria, checked with a plain-text extraction and the portal's own preview, so two graders reach the same verdict.
Greenhouse parses resumes up to 2.5 MB but accepts uploads up to 100 MB, so an upload confirmation proves nothing about whether your fields were read.
Grade each live application pass, fix, explain, or fail on six identity-check criteria, so two graders reach the same verdict before the next screen is booked.
Greenhouse compares three application fields against the CLEAR identity result (last name, email address, and phone number), and a mismatch in any one of them can fail the verification.
Grade any posting's pay line pass or fail on six criteria and reach a comparable annual base figure that a second grader would land within 5 percent of.
A posted pay line is only rankable when six things are recorded: its basis, its period converted at the right hours, its employment-type adjustment, its variable pay split out, its location tier, and whether each of those was stated or inferred.
Match any posting to its jurisdiction row in about a minute and say whether pay was required, whether the range passes, when to ask, and which recourse clock is running.
Washington covers any posting that could be filled by a Washington-based employee, including remote roles, and a line saying Washington applicants will not be accepted does not remove the obligation.
Name any interview request's mechanism, absolute deadline, booking status, and self-reschedule cutoff, then log it, in two minutes without emailing the recruiter.
Submitting a Greenhouse availability grid books nothing: it emails the person who sent the request, and until a calendar invite arrives you have no interview.
Turn an exhibitor list into a tiered, pre-filed booth plan, end every conversation on a named next step, and score each fair by interviews per conversation.
The widely quoted claim that 45% of job fair attendees get an interview comes from NACE's 2024 survey of 20,482 college students, and the report's own Figure 13 puts it at 43.8%, so it says nothing direct about experienced workers.
Set a weekly hours cap and quota, send each qualifying posting within 96 hours, split sends into two tailoring tiers, and reset the mix in a 30-minute Friday review.
Huntr's tracker data puts the sweet spot for an offer at 20 to 39 total applications, and two-thirds of successful searches end within 50, so a quota should be paced for months rather than maximised for a week.
Turn a mixed inbox of scheduling links and availability requests into a capped, buffered interview week with no collisions and no invite unanswered past a business day.
Cap live interviews at 3 to 5 per week, the range the Optimality newsletter recommends, because more than five forces two interviews into one day and splits your preparation.
Score any posting with no pay figure on five 0 to 2 dimensions and land on Apply, Ask First, or Skip, with a written reason a second reader would reach from the same inputs.
A missing salary range on a senior role is weak evidence of low pay, because research on 141M job posts found high-wage firms are the more opaque ones.
Score any lukewarm interview invite on six dimensions and leave with Book, Rehearse, or Decline, a prep cap, a slot date, and a reply sent within a business day.
Any failed dealbreaker (pay, work format, location, or level) forces Decline, whatever the other five dimensions score.
Take one base resume through a queue of postings and log, for each, must-have coverage before and after, the exact edits, minutes spent, and an overfit pass.
In Huntr's self-reported data, tailored resumes drew interviews at 1.6x to 2.15x the rate of untailored ones depending on the report, so treat the lift as a range and not as a promise.
Score one employer's matching roles on six dimensions and write a Pick One, Cluster, or Route verdict with a send order and a one-line note.
Google is the only large employer with a confirmed published cap: up to three applications in a rolling 30-day window, and 90 days before reapplying to the same job.
Turn one master resume into two to four routed variants with locked fields, frozen sent files, per-employer routing, and a weekly sync, so no copy contradicts another.
Greenhouse flags a candidate as a duplicate if any one of email, phone, or LinkedIn URL matches an existing record, so every resume variant you send to one employer ends up on the same record.
Classify each search activity from your week as a direct contact, secondary activity, or nothing, and know which log fields and saved proof it needs before you certify.
Ohio treats several roles at one employer in the same week as one application, according to an Ohio attorney's summary of agency practice, while Virginia, Connecticut and New Mexico count each distinct opening, so a tracker's application total can overstate your countable contacts.
Write a query set for each job source, collapse duplicates by requisition ID, and measure your feed's weekly precision and miss rate on your own target role.
In Refolk's index, US Product Owner profiles (14,533) are 17.7% of the combined Product Manager plus Product Owner pool, so a query on "product manager" alone misses roughly a fifth of the market by construction.
Every stage from the recruiter screen to the final round, and what each one is actually testing.
Take one recorded-interview invite from rules screen to submit, with answers sized to 70 to 80% of the limit, a written retake rule, a transcript check, and a dated log.
In a large vendor dataset, the average one-way interview answer used 70.2% of the time allowed, and the median used 75%, so size rehearsed answers to roughly 70 to 80% of the stated limit.
Grade any recruiter thread against seven written criteria, reach the same Go, Hold, or Withdraw verdict a second grader would, and send the recap that fixes the gap.
A loop is only Go when seven criteria hold in writing: a role-specific range, one named level, a stated location basis, base and total labelled separately, a written source, your floor at or below the range midpoint, and a stated down-level policy.
Map each hiring step to its barrier, send a low-disclosure request to the right contact, and hold written confirmation or a logged escalation before the link opens.
HackerRank's accommodation time only applies if the employer adds it before the candidate starts the test, so the request has to be confirmed before you click start.
Turn your invite pack into a competency-by-exercise grid, drill each exercise to its real clock, and run the day with moves an assessor can score.
Assessors rate each exercise independently and combine the ratings only at the wash-up, so a 7 in one exercise does not rescue weak scores elsewhere. On the Civil Service 1 to 7 scale, a single 3 is allowed only when every other score is high.
Grade every figure on your sent resume pass or fail against seven criteria and leave with a one-page derivation sheet marking each failure restate, range, or retract.
A resume figure is interview-ready only when it passes seven criteria: named baseline, comparison period, measurement source, own share, cross-document consistency, disclosure-safe phrasing, and a spoken derivation of 60 seconds or less.
Grade any take-home brief on eight fixed criteria, reach the same Accept, Rescope, or Decline verdict a second reader would, and send a per-criterion reply.
Plan every take-home at twice its stated hours plus setup time: a brief at the 3-hour published median costs about 6 hours, which is above every unpaid cap in the five sources this standard draws on.
Size how far below band your stated number sits, pick a reset trigger, send a dated written revision before approval, and classify the offer that comes back.
In buyer-seller negotiation studies, the first number explained up to 86% of the variation in final prices, but no study has measured this for a candidate's screen answer, so treat a low number as a risk, not a sentence.
Pull the resume an employer holds, diff it against your master, predict 15 to 30 questions from its tailored lines, and log a hit rate for every round.
Prep from the file the employer holds, not your master: Workday documents no candidate download and no edits after submission, so your tracker's resume-version column is often the only reliable copy.
Sort every lost final round as controllable or not, check your conversion against a named denominator, and name the one round type to change before your next loop.
Published final-round benchmarks disagree by about 15 points because Gem's 39% counts onsites held while Ashby's roughly 24% counts candidates passing the stage, so always write the denominator next to the benchmark.
For any pay question in a screen or loop, find its row and name what it collects, whether a ban covers it, which exception applies, and the reply that protects your number.
Every interview pay question collects one of six things (history, expectations, structure, competing offers, productivity, or level), and the reply class you pick should follow from which one it is.
Score any team-built resume line on six dimensions, get a Claim, Split, or Scope call with its opening sentence, and know when the resume itself needs fixing.
Every team-built resume line can be scored 0 to 2 on six dimensions (decision ownership, execution share, follow-up depth, verification alignment, scorer format, and employer emphasis), and the total out of 12 gives a Claim, Split, or Scope call.
Turn your own role-play brief into a prospect sheet, question plan, timed agenda and objection list, then score a taped run on the weights panels use.
Published sales role-play rubrics agree on five to six dimensions, and next step alone carries 20% of the score in two of them, so a call that runs out of time before the close gives up a fifth of the marks.
Grade any finished take-home against seven pass or fail gates, and know for each failure whether to fix it before sending or name it in the cover message.
A take-home is ready to send only when it passes all seven gates, because Medium's published rubric treats a Mixed grade as more No than Yes, so a half-met criterion tends to grade as a fail.
Find the row for any follow-up asked about a resume line, name the probe, and have the two or three facts that pass it ready before the call.
Any number on a resume triggers a fixed set of five metric probes: the baseline, the time period, how it was measured, what else changed at the same time, and whether the result repeated.
Go from the round invite and your resume to one scored project choice, a timed opening, a probe-answer sheet, and a hostile mock, with every number matching your resume.
Published deep dive formats run from a 10 to 20 minute intro-call version at Ashby to 60 minutes at Hiro, so confirm length, audience, slides, and project reuse with the recruiter before preparing anything.
Score one down-level signal on six dimensions and leave with an Accept, Contest, or Walk verdict and the exact reply to send before the next round.
Published midpoint progression between adjacent professional grades is 10 to 15%, but adjacent-level median total pay gaps at five US tech ladders run 24 to 43%, so price a down-level on total compensation, never on base salary.
Take your own dataset take-home from brief to panel with a dated clarifying email, a capped hour log, an 8 to 12 slide deck, and ten rehearsed Q&A answers.
Company hour estimates for take-homes run low: EasyApply tells employers that if a task takes them 3 hours, it will take candidates 4 to 6.
Take any timed assessment invite through seven dated stages and finish with a written record of the deadline, proctoring setup, retest wait, and result validity.
USA Hire reuses a completed assessment for 12 months from the completion date and allows no retest in that window, so one weak sitting counts against every application in the same series that year.
Name any assessment's test type and scoring model from the invite, judge whether prep will pay, spot the active proctoring, and time an adjustment request correctly.
Situational judgement tests are scored against a key built from subject matter experts' judgements of the best and worst options, and commercial coaching raised SJT scores by d = 0.59 in a high-stakes setting.
Match any mid-loop recruiter update or silence to its likely hiring states, send one question that separates them, and set a dated Live, At risk, or Parked status.
In Metaview's data on 811,298 scorecards, the median was submitted 2.3 hours after the interview and 82.1% arrived within 48 hours, so "waiting on feedback" after 48 hours usually means one or two panelists are missing, not that the decision is pending.
Reading an offer, comparing two, and negotiating without torching the relationship.
Turn an eApp invite into a signed, released questionnaire with prints logged inside the 14-day hold, every entry matched to your records and the resume you were hired on.
DCSA may hold an eApp investigation request for up to 14 days waiting for fingerprints, then flags it unacceptable and returns it to the submitting office.
Grade what you know about your future manager on eight pass or fail criteria and reach a Pass, Hold, or Fail verdict that a second grader would match.
A future-manager file passes only when it holds at least three sources who worked with the manager in the last three years, two of them independent of the employer and one a former direct report you found yourself.
Score one written offer's title on six dimensions and leave with a call to accept, retitle, or dual-title, the request wording, and the confirmation it must produce.
The Work Number returns one job title field, filled from payroll data that employers send weekly, so the HR title you sign for is the string future background checks are most likely to confirm.
Turn a conditional offer email into pulled records, a per-employer reconciliation, a screening form that matches them, and a proof pack within five business days.
The Work Number's Employment Data Report is free and instant on the web, covers dates of employment, title, salary history and pay frequency, and lists every verifier who requested your data in the past 24 months.
Look up any status on your background check report and say what it proves, who acts next, which document clears it, and how long before a final decision.
Clear, Eligible, and Meets company standards mean your results fit the employer's own rules, not that you are hired, so keep other offers alive until the employer confirms in writing.
Turn a reference request into a dated run sheet that clears the check without exposing your search to your current boss or stalling the offer.
Pick three referees, at least two of whom managed you, plus a backup: SkillSurvey requires at least three references with two managers and cannot produce a report from one.
Lock your weights first, grade every score by its evidence, check the workload and manager floors, and reach Take, Close, or Probe before the earlier deadline.
Lock your dimension weights, comp-tie threshold, and close band on a dated sheet before you read the second offer's details, or the offer you already prefer gets to set its own weights.
Turn your interview notes into a claim ledger, test every claim on post-offer calls, and leave each one confirmed, contradicted, or unanswered with a matching move.
In the 2013 Glassdoor/Harris survey of 2,054 adults, 61% said parts of a new job differed from what the hiring process led them to expect. Morale, job responsibilities, work hours, and the supervisor's personality were cited most.
Sort your own flagged verification line into a source error, a resume error, or a structural mismatch, then send the right response to the right party before the deadline.
A Discrepancy label means the screener reached a record and found a field that differs from your entry, while Unable to Verify means its standard channels failed, not that it found you never worked there.
Grade any offer letter against your call notes, sort each gap as drafting error, policy, or reopened negotiation, and name the written fix each Hold needs before you sign.
Grade each agreed term in an offer letter as Match, Query, or Hold against notes you dated on the day of the call, and do not sign while any Hold is open.
Sort every pre-hire project into list, leave off, or ask first, write exhibit lines that reveal nothing, and get a countersigned copy filed before day one.
A blank prior-inventions exhibit is not neutral: on the Zillow form, attaching no list means you represent that there are no prior inventions, and Mavenir's form says the same if the exhibit is unsigned.
After each hiring-manager call, score the team on six weighted dimensions and send the recruiter a take, hold-out, or widen reply within 24 hours.
Written approval validity is counted in months (about 12 at Google and 6 at Amazon, per candidate reports), but recruiters actively push a packet for only about 2 to 8 weeks, so score your window from the last time your packet was sent.
Grade any planned LinkedIn or resume change after signing an offer as Pass, Hold, or Fail, and reach the same grade a second grader would.
A profile or resume update after signing passes only when every gate for that edit type has cleared in writing and every field matches the signed offer letter. If a gate is still open the grade is Hold, and if the edit is already live with a gate open it is Fail.
Name any call booked after your final round, tell whether it still grades you and what it can change, and walk in with three questions and a list of topics to avoid.
Whether a team-match call is graded depends on the order of match and committee: at Google L5 and above the match comes before the committee and feeds the hire decision, while at Meta the committee fixes hire and level first.
Split any retention counteroffer into five parts, score each 0 to 3, and reach a Stay, Firm Up, or Leave verdict with the exact terms to request in writing.
The often-quoted claim that 80% of people who accept a counteroffer leave within six months has never been traced to a published study, so it should not drive your decision.
Grade every onboarding packet item pass, hold, or fail, from verifying the portal through I-9 timing, arbitration and invention clauses, to every benefit deadline.
A fake onboarding portal asks for the same SSN, bank, and ID data as a real one, so the only reliable test is a callback to HR on a phone number you found yourself, never one from the recruiter's email.
Score an offer with a deadline against every open loop on five dimensions and reach accept, a sized extension request, or decline, against a written threshold.
A 3 to 7 day offer extension covers roughly one hiring stage, matching the reported 6-day median from final interview to offer, so only loops already past their final round fit an extension request.
Match each clause in your offer letter to a row and state what it tests, how long it takes, what the employer must do before withdrawing, and the written proof it cleared.
Most individual pre-employment checks clear in 1 to 5 business days after you sign the authorization, and employment verification, not the criminal search, usually sets the finish date.
Turn a signed offer into one dated calendar and pick a notice date and last day that keep every vest, payout, and coverage day, or price the gap to ask for.
Do not resign until you have written confirmation that the new employer's background check adjudication is complete, because a dispute can pause the decision for 30 days plus a 15-day extension and the employer does not have to hold the role open.
Career changes, layoffs, gaps, re-entry, and the searches that do not go to plan.
Confirm transfer eligibility, time the manager talk to your policy's trigger, apply inside the internal window, and price the move against the new role's range.
In a sample of six published transfer policies, five set the minimum tenure at six months, but some count it from hire date and others from time in the current role, so a recent promotion can reset the clock.
Set your filing date around severance, match the separation reason, audit every wage quarter, certify the waiting week, and name the day your first payment is late.
Under the federal first-payment standard, a first unemployment payment is late 14 days after the week-ending date of the first compensable week in waiting-week states, and 21 days after it in states with no waiting week.
Grade your post-layoff setup on eight criteria with evidence attached, and know which of the five blocking Fails must clear before any application goes out.
After a layoff, five of eight launch criteria block the first application on a Fail: records, separation terms, the unemployment claim and work-search rule, the departure line, and the written target set.
Turn an old-field resume into a dated hybrid where every bullet cites a target posting and two target-field strangers name the new role from the top third.
In Refolk's index, only 15 of 12,419 current US Instructional Designers (0.12%) keep "former teacher" in their headline, so people who have already switched lead with the target title, not the old one.
Grade your career-break answer and its two follow-ups on seven pass or fail criteria that a second grader, reading the same transcript, would score the same way.
A spoken career-break answer is finished only when a timed recording and its transcript pass all seven criteria: length, date match, disclosure boundary, current-skill proof, role tie-in, probe readiness, and tone markers.
Cut your manager resume to an IC draft with every bullet tagged and decided, one target level per company, verbatim titles, and a dated recency line per core skill.
Of 8 secondary level-mapping pages, 6 place a first-line engineering manager at the same level as Staff, 1 places the role at Senior, and 1 ranks it above Staff and Principal, so read the target company's own ladder before choosing a level.
Grade every gap on your resume pass or fail against seven criteria, from the 8-month threshold to your employers' payroll dates, and fix each failure before you apply.
In a US field experiment of about 12,000 resumes, callbacks at 8 months unemployed were about 45% lower than at 1 month, and most of that drop happened by month 8, so any gap of 8 months or more needs a label entry.
Match every course, license, certification, or side job to its row and write a resume line with the right section, status label, and proof a verifier can confirm.
In Refolk's index of professional profiles, US headlines containing "career break" outnumber those containing "returnship" 36.4x (2,655 to 73), so most returners have to prove recency line by line rather than through a formal program.
Map what every former-employer channel will disclose, then write and rehearse one termination answer that no record, reference, or form contradicts.
The dominant automated verification service confirms dates, title, and pay but cannot provide reasons for separation, so the firing usually surfaces through the separate HR rehire question, not the database.
Turn your weekly benefit, your state's partial formula, and a gig's hours and pay into a week-by-week ledger that flags the exact week the gig zeroes out your claim.
On a $275 weekly benefit and a $200 gig week, the partial benefit ranges from $125 in California to $213 in Illinois, because each state measures its earnings disregard from a different base.
Clear a four-part go or no-go gate, publish one post with one target role and one ask, and log every useful reaction with an owner and next action by day 14.
In Ashby's analysis of nearly 250,000 hires, referrals were about 1% of applications but about 18% of hires, so a layoff post should ask for introductions, not applications.
Grade any conviction-question answer pass or fail on six fixed gates, so two people checking the same answer against your record reach the same verdict.
A conviction-question answer passes only if it clears all six gates: scope parsed, record verified, sealing handled, jurisdiction timing checked, explanation shaped, and consistency with the later background check.
Fill every employment-history field after a layoff so it matches your separation paperwork and a verification record, with no knockout checkbox and no later contradiction.
An automated employment verification such as The Work Number returns hire date, termination date, total time, job title and rate of pay, so the end date and title on your form have to match your separation agreement exactly.
Score six dimensions and leave with one verdict (elect COBRA, enroll in a marketplace plan, or hold the election) plus a dated action list and a last-safe-day.
COBRA's jump in price comes from a change of payer, not a price rise: on KFF's average single premium of $9,804, an active worker pays about $139 a month while COBRA costs about $833, which is 6.0x as much.
Total a credited-months figure for any stated minimum, one source at a time, and enter a years-of-experience answer that survives a date-and-hours check.
Under federal OPM rules, 20 hours a week for 12 months is credited as 6 months of experience, because credit is prorated against a 35 to 40 hour week.
Build a one-page separation ledger, source every line, and grade your layoff packet ready to sign, ready to query, or blocked the way a second reviewer would.
Every exit payment belongs in one of two columns, owed regardless or conditioned on the release, and a conditioned line that duplicates an owed line fails the federal age-waiver test for consideration "in addition" to existing entitlements.
Hear any version of why you left or what you have been doing, find its row, and give a short answer that matches what an employer can verify and the law allows.
Unemployment-status laws in New Jersey and Oregon cover job ads only, and none of the jurisdictions with these laws bans asking why you left in an interview.
Turn your last day and I-94 date into a dated calendar with an offer deadline, LCA and petition buffers, and a fixed trigger day for a fallback I-539.
At a published university pace of about two weeks each for case initiation, the LCA, and petition prep, roughly 42 of the 60 grace-period days go to paperwork, leaving about 18 days to search.
From one posting and your history, compute countable years, score five dimensions, and land on lateral, step down, entry, or bridge, plus whether a second level is safe.
Under the US federal crediting rules, a year at 20 hours a week counts as 6 months, so three part-time years at that pace are 1.5 countable years, and that is the number a screening question should get.
Turn your application history into stage conversions per cluster and channel, name the one stage that is failing, and test a single fix in four weeks.
In the Kroft, Lange and Notowidigdo audit study, most of the callback decline from a longer unemployment spell happens in the first eight months, after which callbacks flatten near 4%, so at month six the gap is mostly a sunk cost and target choice is the lever left.
Where the hiring actually is: by role, by city, by industry, and what the numbers mean for your search.
Given one employer, your SOC code, metro, and offer, you will score six dimensions from public filings and reach a Target, Probe, or Drop call a second reader would match.
Under wage-weighted H-1B selection, a Level IV registration gets four lottery entries and a Level I registration gets one, and DHS modelled Level I selection at 15.29% against 30.58% for Level II.
Build a ranked list of about 30 employers posting your role family now, each with live counts, a fit sentence, and a contact path, refreshed in 15 minutes a week.
No published target-list method ties list size to measured outcomes: recommendations run from 15 to 20 up to 50 or more, and every one is based on manageability or resilience.
In about three and a half hours, turn 30 deduplicated postings for one role into a ranked demand table and a base resume ordered by it, with every gap logged.
At 10 postings, an item that 7 of them ask for has a 95% interval of roughly 40% to 89%, so 30 unique postings is the smallest sample where a 70% share reliably means most of the market.
Grade any interview review, forum post, or prep guide Admit, Hold, or Reject on seven written criteria, and reach the same verdict a second grader would.
Glassdoor does not remove or archive interview reviews because of their age and does not fact-check them, so the expiry date and the accuracy check have to come from the reader.
Turn any US venture-backed employer's Form D filings and headcount into a low, mid, and high months-left range and a raise-by month you can check against the founder.
For any one offering, the cash raised is the amount sold on its latest Form D amendment, because amendments restate cumulative totals and the SEC itself counts only the increments.
Set up two job targets as matched arms, size them from your own interview rate, send on a controlled schedule, and stop on a written rule with a dated shift, split, or hold.
A conventional test needs about 20 interviews in the weaker arm, which means 432 applications per arm at a 5 percent baseline to detect a doubling, and 1,138 per arm at a 2 percent baseline.
Grade every line of any posting as Gate, Screen, or Wish with written tests, and decide apply or skip from the Gates you clear, not a match percentage.
A requirement only becomes a Gate when outside evidence backs it: a law, a license, work authorization, or a statutory age, never the word Required in a header.
Starting from one held title, you will produce five to eight adjacent titles, each with an overlap score, level match, same-day posting count, and named gap.
O*NET gives every occupation exactly 20 related occupations in three tiers, but the tier order comes from expert review rather than a published overlap score, and 35.1% of the top-five links cross into a different SOC major group.
Sort your own postings into four degree labels, write an education block that never reads as a finished degree, and make a send, route, or drop call for each one.
Across 11,300 roles studied by Harvard Business School and the Burning Glass Institute, dropping the degree requirement raised non-degree hiring by only 3.5 percentage points on average.
Score every interview round type across 10 to 30 target employers and leave with a weekly hour split labeled Drill, Refresh, or Skip, plus a dated re-score.
Score interview prep per employer-by-round pair, not per employer, because one company can run an AI-enabled coding round with its own provided tool while banning outside AI in its other rounds, as Meta does.
Start from any one sighting, match every copy of a role to one requisition ID, flag stale and agency copies, and log a single application channel.
Workday requisition IDs stay the same when a role is reposted, so a shared ID proves two copies are the same job but does not prove the copy in front of you is still open.
Grade every term on your resume against 20 dated postings with written count thresholds, so a second grader reaches the same verdict on every term.
Grade resume terms against 20 target postings pulled on one date, because two postings for the same job title share a median of only 25% of their named requirements.
For any in-demand skills claim you meet, name what it counts, its window, and whether it can justify adding, moving up, or cutting that skill on your resume.
O*NET flags a skill as In Demand for an occupation when it appears in 5% or more of that occupation's US employer postings, so a flagged tool can still be missing from 95% of those postings.
In under an hour, map any employer's interview loop: assessed rounds, days from first interview to offer, a take-home flag, a confidence level, and a same-clock benchmark.
On Glassdoor's own employer page, the company-wide interview average is 24 days on 1,301 reports, but software engineers report 12 days on 27 reports, so the headline number is twice the role-level one.
Take any published hiring number, name its clock, unit, and sample, and turn it into a figure you can compare with your own loop, or reject it.
Time to fill starts when the employer approves the requisition and time to hire starts at the candidate's application, so SHRM's 39-day nonexecutive median includes days before anyone applied. An older SHRM report puts that pre-screening phase at roughly 12 days (derived).
Look up any job board label, name who set it, judge what it proves about competition, and pick apply now, apply via careers page, verify first, or ignore.
On Indeed, Urgently Hiring comes only with premium sponsored tiers, and Indeed's own figure is 74% more applications than sponsored jobs without it, so the label signals more competition, not less.
For every past role, score your official title on five dimensions and land on keep, pair, or translate, with a written reason and a verification note ready.
Job titles explain more than 90% of the variance in posted wages, against about a third for six-digit SOC codes, so an off-market title throws away the main signal a recruiter uses to price you.
Trace any listing to its ATS record and real first-published date in under two minutes, then sort and send the day's queue by true age.
A job board's "posted X ago" counter dates the board's copy of a listing, not the employer's original requisition, so late copies, promoted renewals, and reposts all make old jobs look new.
Match any posting's location pay wording to one of six named methods and work out what percentage of anchor-market pay to expect in your own metro.
Pave's salary data puts Tier 2 base pay at about 89% of the anchor market and Tier 3 at about 84%, but its new-hire equity data puts the same tiers at 71% and 64%, so the equity discount is roughly twice the base discount.
Take any certificate or bootcamp from public outcome records to an expected break-even month, a downside month, and a no-placement case you can defend before you enroll.
Only about 1 in 8 non-degree credentials produces wage gains of 10% or more that graduates would not otherwise have had, so a break-even month built on a provider's advertised salary is a best case, not a forecast.
Put this to work
Drop a resume or a LinkedIn URL. I rank the live openings against it, rewrite the resume and write a cover letter for the best of them, and fill in the employer's form when you press the button. You read, you decide what goes out.
01Drop your resume
A PDF or a LinkedIn URL. About a minute, once.
02I rank the openings
Every weekday morning, the live catalog scored against your history. Up to 20 worth your time, not two hundred links.
03Each one is written up
Resume rewritten for the posting, a cover letter, a fit score. Press send, or let me fill in the form.
Free, no card. Up to 25 applications sent a week, and nothing is sent until you say so.
Every guide is also served as plain Markdown at /candidates/guides/<slug>.md, and the index at /candidates/guides/index.md.