Why Good People Can’t Get Jobs

1. 1Calculate the cost of keeping each vacancy open before treating an unfilled job as free: count overtime, burnout risk to remaining staff, delayed projects, and lost revenue attributable to the role — not just the salary you're saving.
2. 2Separate real qualifying requirements from wish-list items in every requisition. Force the hiring manager to label each skill 'deal breaker' or 'nice to have,' and challenge every deal breaker that isn't legally or technically mandatory.
3. 3Run the hurdle math on any requisition: multiply the pass rate for each requirement together. If ten requirements each screen out half of applicants, only about 1 in 1,000 qualified people survive — flag any requisition that will mathematically produce zero hits.
4. 4Frame each open role as an explicit make-versus-buy decision. Compare the cost and time to hire experienced talent externally against training an internal or lower-skilled candidate to fill the same job.
5. 5Use Bidwell's evidence to price the two options honestly: outsiders take three years to match insider performance and are paid more, while insiders take seven years to catch the outsiders' pay — internal development is cheaper and faster to competence but slower to reward.
6. 6When labor is scarce, lower requirements rather than leaving the seat empty. Recognize that the same job can be done many ways, so a candidate who lacks a listed tool can still perform the tasks.
7. 7Fix HR's incentive structure so the function is measured on filled roles and time-to-fill, not blamed solely when a hire fails — otherwise it will never push back on inflated requisitions.
8. 8Audit your own screen by applying anonymously to your own openings, the way the Philadelphia-area HR executive did, to see whether qualified people are being rejected.

A Philadelphia-area HR executive suspected his company's hiring standards were too high, so he applied anonymously for a job in his own firm — and did not make it through the screening process. This is HR analytical capability in miniature: instead of accepting the 'we can't find qualified people' story, he ran a direct test of the screen itself and found the process, not the labor market, was the constraint. Contrast this with the manager Kim Hogan describes, whose single post — demanding one person be the HR, marketing, publishing, project-management, accounting, and finance expert — sat unfilled for nearly a year because the employer insisted on the 'right fit.' A capable HR function would have flagged that requisition as a 'unicorn,' run the hurdle math showing near-zero qualified applicants, and pushed the hiring manager to separate deal breakers from wish-list skills. The Fernando Boero listing — a mechanical engineer role requiring 65 words-per-minute typing — shows the same failure: nobody was left to ask whether that requirement was a deal breaker. The tell in each case is a long-open vacancy paired with a requirement no one can defend on business grounds.

• If the requisition lists more than a handful of deal-breaker requirements, then run the hurdle multiplication and reject the requisition when the expected qualified pool approaches zero.
• If a single role bundles the work of three prior jobs plus computer skills 'because we can,' then split the role or drop non-essential requirements before posting.
• If it costs more to keep the vacancy open (overtime, burnout, stalled projects) than to hire a lower-skilled candidate and train them, then fill and train.
• If everyone in your industry is trying to 'buy' the same experienced skills from competitors, then shift toward 'make' — training internally — because the buy strategy fails collectively even when it works for one firm.
• If HR is held responsible when a hire fails but not rewarded for filling roles, then don't expect pushback on inflated requisitions — change the incentive first.

• Treating an unfilled vacancy as a cost saving because short-term costs drop without revenue falling — the tell is a role reported as improving profitability while projects stall and staff burn out.
• Writing vague job descriptions because hiring managers can't agree on what they want — the tell is a flood of applications and a role that stays open for months.
• Piling every conceivable skill into a requisition to minimize the risk of a bad hire ('looking for a unicorn') — the tell is a listing demanding one person cover five unrelated functions.
• Defaulting to 'buy' over 'make' because outside hiring feels easy — the tell is no training budget and a belief that vacancies must be filled with someone who can 'step right in.'
• Automating the entire screen to cut HR head count so no human is left to challenge requirements — the tell is qualified candidates, including your own executive, silently filtered out by parsing software.

• ▸Role & tasks: list what the job actually does, not the ideal résumé
• ▸Requirements, one per line: mark each 'deal breaker' or 'nice to have'
• ▸Deal-breaker justification: for each, why it's legally/technically mandatory
• ▸Hurdle math: pass-rate estimate per requirement, multiplied — expected qualified pool
• ▸Vacancy cost per month: overtime + burnout risk + delayed revenue/projects
• ▸Buy option: cost and weeks to hire experienced talent externally
• ▸Make option: cost and weeks to train an internal/lower-skilled candidate to competence
• ▸Decision: fill-as-is / lower requirements / train internally — with the deciding number
• ▸Owner accountability: who is measured on filled role and time-to-fill

1. 1Diagnose search intensity as an economic choice, not a candidate deficiency: ask whether you are waiting because no one qualifies, or because so many qualified applicants are arriving that holding out feels cheap.
2. 2Track applicant volume per vacancy over time. When the ratio spikes (recessionary conditions bring you thousands of applicants), recognize that your recruiting effort per vacancy has likely fallen—that decline is the search intensity signal.
3. 3Separate 'qualified' from 'perfect.' Write down the tasks the role must perform, then list which of your requirements are must-haves versus preferences you added because pickings are good.
4. 4Make the cost of the unfilled vacancy visible: quantify overtime, burnout of existing staff, and delayed projects, so the invisible cost of waiting competes with the visible cost of hiring.
5. 5Test whether your 'shortage' is really a price problem: ask if a qualified candidate would take the job at the wage offered, and whether you are holding out for someone who will work below market rate.
6. 6Set a stopping rule before you start: decide in advance the point at which variation among candidates no longer justifies continued searching, and hire the qualified applicant rather than chasing the overqualified one.
7. 7Audit whether relative competition is silently raising your bar—if you keep passing over qualified people because someone overqualified applied, name that behavior and its consequences (older and perfectly-qualified workers screened out).

An employer reported 25,000 applicants for a standard engineering position, and the staffing people declared not one was qualified. Cappelli treats this as the tell of excess search intensity, not a skills gap: with 25,000 people arriving at no cost, searching is nearly free, so the firm holds out for perfect. The 'perfect not qualified' logic here is the prom-date dynamic—so dazzled by choices that you put off asking anyone. Meanwhile the mechanics compound it: applicant-tracking software screens on keywords, so one reader was told he was perfect except his prior job title didn't match a title unique to that company. The point is that a vacancy sitting open amid thousands of applicants is evidence of cheap search, not proof that something is wrong with the candidates. Pundits have said the opposite after every recent recession and, Cappelli notes, are always wrong.

• If applicants are plentiful and cheap to attract, then assume your requirements have crept upward and re-examine them before concluding no one qualifies.
• If you find little variation in candidate price and attributes, then stop searching and hire; if you find lots of variation, searching longer may pay—but bound it with a deadline.
• If you're passing over a qualified applicant only because someone overqualified appeared, then ask what the overqualification actually buys you before rejecting the qualified one.
• If qualified candidates decline at your wage, then treat it as a price problem ('won't pay the market wage'), not a shortage—raise the wage or accept the tradeoff.
• If you can't find someone with the exact skills, then invoke the make-or-buy choice: hire basic ability and train, rather than waiting indefinitely for a precise fit.

• Reading an open vacancy as proof candidates are deficient—the tell is you cite the vacancy itself as evidence, ignoring that cheap search alone prolongs it.
• Letting requirements silently ratchet up because applicants are plentiful—the tell is job specs that grew stricter as unemployment rose, with no new task actually added.
• Waiting for 'perfect' while blind to the cost of waiting—the tell is existing staff burning out and projects stalling while you keep the seat open.
• Mislabeling a wage gap as a skills gap—the tell is 'we can't find anyone' when the truth is 'we won't pay the market wage,' as with diamonds you can always buy at the price.
• Trusting keyword screening as if it understands applications—the tell is rejecting a 'perfect' candidate over a nonmatching, company-unique job title.

• ▸Vacancy title & unfilled duration: how long has it been open?
• ▸Applicants received vs. hires made: is the ratio high and rising?
• ▸Must-have tasks (not preferences): what must the person actually do?
• ▸Requirements added since applicants got plentiful: which bars crept up?
• ▸Wage offered vs. estimated market wage: is this a price problem?
• ▸Cost of leaving it open: overtime, burnout, delayed projects (quantified).
• ▸Make-or-buy option: could a basic-ability hire be trained instead?
• ▸Screening filter risk: any keyword/title mismatch rejecting good fits?
• ▸Stopping rule: at what point do you hire qualified over perfect?

1. 1Count the actual hurdles in your requisition. List every requirement built into the screening software and treat each as an independent probability gate. If a typical applicant clears each at 50%, ten hurdles pass roughly 1 in 1,000, and fourteen produce the '25,000 applicants, none qualified' outcome.
2. 2Separate deal-breakers from preferences. For each line, ask the hiring manager the Cappelli test: 'Is this a deal breaker, or would you hire someone who lacked it?' The mechanical-engineer req demanding 65 words-per-minute typing is the tell that nobody scrubbed the list.
3. 3Audit the software for exact-match traps. Key-word and job-title matches reject capable people whose prior title differs from your company-unique title. Add synonym sets and demote title-matching from pass/fail to a scored factor.
4. 4Distinguish a skills shortage from a wage decision. If your only complaint is 'qualified candidates won't take the wage offered,' you don't have a shortage — you have a price you won't pay. Diamonds are expensive but not scarce.
5. 5Run a make-or-buy check before rejecting. When precise skills are rare or costly, hire basic ability plus ramp-up/training. During the 1990s IT boom, only ~10% of people in real IT jobs held IT academic credentials because the credentialed were too expensive to insist on.
6. 6Test your own screen. Have an insider or your best current performer apply anonymously; if they fail, your gates are broken, not the labor market.
7. 7Handle near-misses legally and deliberately. Pulling rejected applicants back for a second look risks adverse-impact challenges because it treats applicants inconsistently — so fix the screen upfront rather than fishing survivors out afterward.

A job seeker was told the temp already in the role was doing all the duties, including the added ones, and doing them well. When he asked why they hadn't simply hired her, they said she failed the online questionnaire — mostly a personality-fit test — so they wouldn't hire her. Here the person with demonstrated on-the-job success is screened out by a gate that has nothing to do with ability to perform the work. The screen is measuring the wrong quality (fit questions the software can't accurately assess) and treating it as an absolute hurdle. This is not an isolated tale: 38 percent of employees cited in the passage sit inside processes like this. The fix was available and ignored — someone already knew a capable candidate existed and let the algorithm veto her.

• If a current temp or employee already performs the job well but fails your screen, then the screen is wrong — override it, don't defend it.
• If your only unmet requirement is a title match unique to your company, then convert it to a scored preference, not a pass/fail gate.
• If you have 14+ requirements each treated as mandatory, then expect near-zero qualified applicants regardless of the real talent pool — cut hurdles.
• If candidates decline at the offered wage, then classify it as a pricing choice, not a shortage, and decide whether to pay market or lower requirements.
• If the skill is common and cheap in the market, buy it; if it's rare or firm-specific, train basic-ability hires with ramp-up time.

• Encoding trivial requirements as deal-breakers — tell: a req list where an engineer needs 65 wpm typing, or a cotton-candy operator needs prior cotton-candy-machine experience.
• Trusting title/key-word matching — tell: 'perfect except your previous job title didn't match ours,' rejecting an otherwise ideal candidate.
• Gutting HR so no one challenges the requisition — tell: nobody left to ask 'are all these skills deal breakers?' and it feels safer to leave the job vacant.
• Mistaking a wage gap for a skills gap — tell: 'qualified people won't work at the wages we offer' stated as evidence of shortage.
• Fishing rejected applicants back out ad hoc — tell: reevaluating selected rejects, which treats applicants inconsistently and creates adverse-impact exposure while never fixing the underlying screen.

• ▸Requirement — write each requirement built into the screen, one per line.
• ▸Deal-breaker? (Y/N) — mark N unless a capable performer genuinely could not do the job without it.
• ▸Pass estimate (%) — estimate share of willing applicants who clear this one gate.
• ▸Cumulative pass — multiply pass estimates down the list to see survivors per 1,000.
• ▸Title/key-word trap? — flag any gate that rejects on exact title or keyword rather than capability.
• ▸Wage vs. skill — note whether unfilled means 'can't find' or 'won't pay market price.'
• ▸Make-or-buy — for each rare requirement, record whether basic-ability-plus-training is cheaper than holding out.
• ▸Insider test result — did your best current performer or an anonymous insider pass the screen? (Pass/Fail)
• ▸Action — for each flagged row: delete, downgrade to scored preference, or keep as verified deal-breaker.

1. 1Define when the vacancy actually costs you: name the specific harm (existing staff burning out, a delayed project) rather than assuming any open seat is a loss, since standing vacancies can persist without hurting operations.
2. 2Benchmark your offered wage against the market rate before blaming applicants — pull the Bureau of Labor Statistics average for the exact job and compare it to what you post; a gap of even 30-50% will empty your pipeline.
3. 3Audit your job description for false 'qualifying requirements': strip company-unique job titles, keyword traps, and specs like 'prior experience on this exact machine' that screen out capable people. Test it by applying anonymously to your own posting.
4. 4Run the make-or-buy calculation explicitly: price the cost of finding a fully-skilled hire against hiring someone with basic abilities plus training or ramp-up time, then choose the cheaper path unless the skill is genuinely unavailable in the market.
5. 5Set the search-effort level to the payoff: search harder only when many candidates compete and the difference between them is large; otherwise stop waiting for a perfect match and hire the adequate candidate now.
6. 6Fix the internal constraint that makes training look impossible — if you 'keep looking and looking' because the department is too disorganized to onboard, build a ramp-up program (even a 90-day probation) rather than raising external requirements.
7. 7Recheck your ATS keyword filters against real resumes of people you'd want; loosen matching so a mismatched title or missing keyword doesn't auto-reject a qualified applicant.

Mechanical Devices, a parts supplier, could not fill 40 machinist jobs, and that shortfall was estimated to cut sales by 20%. The story circulated as proof of a skills gap. But the jobs paid $13 per hour while the Bureau of Labor Statistics average for such machinists was more than $19 — nearly 50% higher. Applying the wage benchmark step reveals the vacancy was a pricing problem, not a skills problem: they were shopping for below-market labor. Raising the wage toward $19 would plainly have widened the applicant pool. The lesson is that an unfilled vacancy plus a below-market wage does not equal a shortage — it equals an offer no capable worker will accept.

• If applicants decline your offers or none apply, then raise the wage toward the BLS market rate before concluding there is a skills shortage.
• If the skill you need exists in the outside market and hiring-plus-training is cheaper than a prolonged search, then hire for basic ability and train.
• If the supply of already-skilled candidates is growing and their wages are falling, then stop training and buy the finished skill instead.
• If many employers compete for the same candidates and they differ widely in quality, then invest more in search; if candidates are interchangeable or few, then hire the first adequate one.
• If your posting contains company-unique titles or exact-tool experience requirements, then rewrite them as transferable capabilities so the ATS stops rejecting qualified people.

• Treating hiring like Home Depot — assuming the job must be filled by a perfect-fit part or the business stops. Tell: you leave vacancies open indefinitely rather than covering tasks or training a near-fit.
• Blaming applicants for a wage gap. Tell: you insist 'our wages are high enough' but haven't compared them to the market rate in years (the ~11% who admit wage-refusal understates the real number).
• Writing vague job descriptions because hiring managers can't agree, which floods you with applications you then over-filter. Tell: thousands of applicants and still 'no qualified candidates.'
• Trusting keyword-matching software to understand applications. Tell: a candidate called 'perfect' is rejected because a job title didn't match, or your own HR exec fails his own screen.
• Letting cost-control bias suppress hiring because you track employee costs to the penny but can't measure the value of filling the job. Tell: the vacancy 'looks profitable' short-term while projects quietly stall.

• ▸Vacancy & real cost: name the concrete harm caused by leaving it open (burnout, delayed project, lost sales %).
• ▸Offered wage vs. BLS market wage: list both numbers and the gap %.
• ▸Qualifying requirements audit: flag any company-unique titles, exact-tool experience, or keyword traps to remove.
• ▸Self-application test result: did a qualified insider pass your own ATS screen? (yes/no).
• ▸Make-or-buy cost: cost to find a fully-skilled hire vs. cost to hire basic-ability + train/ramp-up.
• ▸Market supply signal: is the pool of already-skilled candidates growing and their wages falling? (yes/no).
• ▸Search-effort setting: how many employers compete for these candidates and how much do they differ (drives high/low search).
• ▸Internal onboarding readiness: can the department ramp up a near-fit hire? If no, name the fix.
• ▸Decision: raise wage / loosen requirements / hire-and-train / keep searching — and the trigger justifying it.

1. 1Separate the two cost buckets before you count anything: the vacancy cost (business lost while the seat stays empty) and the mismatch cost (talent wasted and unemployment prolonged when your requirements exclude capable people). They have different fixes.
2. 2Pin down when a vacancy actually starts to hurt. Ask: are the vacant job's tasks being absorbed by other staff (burnout risk), or are they simply not getting done (delayed projects, postponed expansion)? The passages note most firms have no idea when unfilled-vacancy pain begins—establish that trigger point explicitly.
3. 3Correct the accounting bias against hiring. Costs per employee are known to the penny; the value each job contributes is not. Before treating a vacancy as 'free savings,' name the specific revenue, project, or capacity that goes missing so the cost stops being invisible.
4. 4Audit the job requisition for 'unicorn' inflation. Line-item every credential and skill and mark each as deal-breaker or nice-to-have. Requirements that fold three former roles into one, or add trivial screens (65 wpm typing for an engineer), manufacture the vacancy that is costing you.
5. 5Recognize that requirements should move with the labor market. When labor is scarce, requirements should fall to avoid bidding up wages; when applicants are plentiful, resist the reflex to raise them until no one qualifies.
6. 6Weigh make-vs-buy for every hard-to-fill role. If no perfect match exists, cost out hiring basic ability plus training against the ongoing loss of the empty seat—the same tasks can be done in different ways (US engineers/unskilled vs. German craftsmen).
7. 7Screen your own screen: have someone apply anonymously (as the Philadelphia HR executive did) to see whether qualified people are being filtered out before a human reads them.
8. 8Track the distributional cost too: overqualified applicants stuck in low roles, unemployed candidates excluded on principle, and older workers screened out. These are society-borne costs of prolonged unemployment and wasted talent, not just company costs.

Ben Venue Laboratories in Ohio sought to fill 100 openings and drew 3,600 applicants—yet judged only 47 qualified, with many failing a basic reading and math test. CEO Thomas J. Murphy expected that in tough times he'd have his pick of people. Read through the lens of cost: the vacancy cost is the 100-or-so seats' worth of production not running or being absorbed by overworked staff; the mismatch cost is 3,553 people the company won't hire, some of whom could do the work with training under a make-vs-buy calculation. The headline reads as a supply-side failure ('25,000 Applicants, No One Qualified'), but the passages show requirements and screens, not worker character, often produce the gap. The society-level cost compounds: prolonged unemployment for excluded applicants and lost business measured in billions across millions of unfilled jobs. The fix is to test whether the qualifying bar itself—not the labor pool—is generating the loss.

• If the vacant job's tasks are being silently absorbed by existing staff, then treat burnout and turnover as a hidden accruing cost, not a saving.
• If a vacancy 'improves' P&L profitability, then suspect the accounting is hiding lost value—apply the reductio: if empty seats help, why not lay everyone off?
• If a role has gone unfilled for months and the requirement list fuses several former jobs or adds trivial skills, then you're hunting a unicorn—cut requirements before you widen the search.
• If labor is scarce and expensive, then lower requirements to avoid bidding up wages; if applicants are plentiful, resist the reflex to raise them.
• If you can't find the exact skill, then price 'buy basic ability + train' against the running cost of the empty seat before insisting on a perfect match.

• Blaming the supply side by default. Tell: leadership repeats 'schools aren't preparing people' while never checking whether its own screens reject capable applicants.
• Reading empty seats as savings. Tell: a manager or unit looks more profitable the longer a job stays vacant, and no one questions the missing revenue.
• Requirement inflation ('unicorn' postings). Tell: a single job asks for HR, marketing, publishing, project-management, accounting, and finance expertise—and stays open a year while the employer insists on 'the right fit.'
• Trivial disqualifiers. Tell: a 30-year mechanical engineer is screened out for lacking 65-words-per-minute typing—criteria unrelated to the actual work.
• No one left to challenge the requisition. Tell: gutted HR departments rubber-stamp the hiring manager's wish list, so 'Are all these deal breakers?' never gets asked.

• ▸Vacant role & days open: name it and count elapsed days to a filled state.
• ▸Where the work is going now: absorbed by others (burnout risk) / not done (which projects delayed?).
• ▸Value at risk: the specific revenue, project, or capacity lost while empty—not the cost saved.
• ▸Requirement list, tagged: mark each skill/credential deal-breaker or nice-to-have.
• ▸Unicorn check: is this role fusing 2+ former jobs or adding trivial screens? (y/n + which).
• ▸Labor-market setting: applicants scarce or plentiful—should requirements fall or hold?
• ▸Make-vs-buy estimate: cost to hire basic ability + train vs. cost of continued vacancy.
• ▸Own-screen test result: would an anonymous qualified insider pass your filter? (y/n).
• ▸Society-borne costs noted: overqualified/underemployed, unemployed-excluded, or older workers screened out.

1. 1Test the surplus claim before accepting it: compute applicants-per-qualified-hire, not applicants-per-posting. Ben Venue drew 3,600 applicants for 100 openings yet found only 47 qualified — the raw abundance masks a self-inflicted screening problem.
2. 2Separate the two sentences employers conflate: 'We can't find anyone' versus 'We won't pay the market wage.' Pull the Bureau of Labor Statistics average wage for the role and compare it to what you offer. Mechanical Devices paid $13/hr against a $19/hr BLS average — a 50% gap that alone explains the empty applicant pool.
3. 3Audit your job description for 'unicorn' inflation: list every required credential and ask which are genuinely needed day one versus piled on to minimize failure risk. Strip requirements that describe an already-trained incumbent rather than a trainable hire.
4. 4Inspect your applicant-tracking software's keyword rules. Identify job-title mismatches and company-unique titles that auto-reject qualified people — the reader who was 'perfect except' his prior title didn't match a title unique to that firm.
5. 5Recognize the relative-competition trap: when applicants outnumber openings, overqualified candidates bump out the merely adequate and you delay filling to chase 'best fit.' Set a search deadline and estimate the cost of the vacancy staying open so you stop waiting for the perfect date.
6. 6Remove the two exclusions abundance tempts you into: screening out the currently unemployed and older workers. Both categories contain exactly the experienced people you claim to lack.
7. 7Decide who pays for the ramp-up. If no candidate arrives pre-trained, either fund training/investment yourself or raise the wage — those are the only market-consistent options.

Ben Venue Laboratories posted 100 openings in Ohio and received 3,600 applications, declaring only 47 qualified, with many failing a basic reading and math test. CEO Thomas Murphy's own words give away the mindset: 'You would think in tough economic times that you would have your pick of people.' The abundance itself created the expectation that a perfect pick should simply appear. But 47 qualified out of 3,600 is a screening and standards result, not proof that qualified people don't exist — it's the same pattern as the company reporting 25,000 applicants for a standard engineering role with, allegedly, not one qualified. The corrective questions are concrete: what wage was offered against the BLS average, what pre-existing experience was demanded that training could supply, and did automated keyword screening discard capable people over title mismatches? Run those checks and the 'shortage' typically resolves into unrealistic requirements, below-market pay, or software error — not a missing workforce.

• If applicants are plentiful but 'qualified' hires are near zero, then suspect your standards and screening filters before blaming the labor pool.
• If you say you can't fill a role, then check your offered wage against the BLS average; a gap of ~50% (like $13 vs $19/hr) is the shortage, not a talent shortage.
• If a candidate is overqualified and available, then hiring them isn't a loss — the overqualified win the relative competition and often do the job better, but don't reject the merely-qualified to keep waiting.
• If a job description reads like it requires an already-trained incumbent, then you're 'looking for a unicorn'; cut requirements not needed on day one.
• If you're auto-excluding the unemployed or older applicants, then you're discarding the exact experience you claim is scarce — stop.

• Blaming schools and academic skills: CEOs cite the education system, but employers doing the actual hiring complain about lack of work experience, not academic skills. Tell: the complaint targets 'training' and 'ramp-up,' not reading and math.
• Treating vacancies as proof candidates are deficient: pundits say unfilled jobs mean bad applicants; they've said this after every recession and are always wrong. Tell: the vacancy persists while the wage stays flat.
• Calling below-market pay a shortage: like insisting there's a diamond shortage because they're expensive. Tell: 'qualified candidates won't take the jobs' at the wage offered.
• Unicorn job descriptions: piling HR, marketing, publishing, project management, accounting and finance into one 'right fit' posting that sits unfilled for nearly a year (Kim Hogan's example). Tell: the requirement list grows every time a candidate falls short.
• Software over-screening: keyword and exact-title matching rejects perfect candidates over a title unique to your company. Tell: 'perfect for the position, except...' rejections.

• ▸Openings vs. total applicants: raw ratio — is abundance real?
• ▸Qualified-per-applicant rate: how many pass screening, and why do the rest fail?
• ▸Offered wage vs. BLS average: state the dollar gap and percentage.
• ▸Day-one-required credentials: list only what's needed immediately, flag the rest as unicorn additions.
• ▸ATS keyword/title rules: identify auto-rejects caused by company-unique titles or mismatches.
• ▸Excluded categories: are unemployed or older applicants being filtered out? Yes/No + justification.
• ▸Vacancy cost: estimated dollar cost per week the role stays open (to counter waiting-for-best-fit).
• ▸Who pays for ramp-up: training investment, raised wage, or accept overqualified — pick one.

1. 1Separate the wish list from the must-haves: for each requirement listed, ask whether the job can physically be done without it. The cotton candy operator ad demanding prior cotton-candy-machine experience is the tell—strip requirements that describe the exact prior job rather than the underlying capability.
2. 2Audit for 'to get the job you must already have the job' Catch-22: flag any requirement demanding a specific prior title or identical role. One reader was rejected because his previous job title didn't match a title unique to that company—that mismatch reveals the requirement is a screening artifact, not a capability need.
3. 3Run the anonymous self-test: have someone apply for your own open role under a blind profile, as the Philadelphia-area HR executive did—he didn't make it through his own company's screen. If a competent insider fails, the bar is set for a unicorn, not a worker.
4. 4Inspect your applicant-tracking software's keyword rules directly. Since screens reject anyone missing exact key words, pull the reject pile and count how many rejected applicants could actually do the job. Remove hyper-specific or company-unique terms that no external candidate would phrase correctly.
5. 5Recalibrate requirements to labor market conditions: requirements rise when applicants are plentiful and fall when labor is scarce. Consciously lower the bar during a glut instead of letting abundance inflate it—recognize that 25,000 applicants and 'not one qualified' signals your standard, not the supply.
6. 6Separate 'can't find' from 'won't pay': before adding requirements, check whether the real issue is wage. Not paying the market price for talent is not a shortage, any more than expensive diamonds are scarce.
7. 7Make the make-or-buy decision explicit: decide which capabilities you will train for versus require pre-loaded. Since work experience can't be gained in school and you're refusing to train or invest, you've engineered your own shortage.

A company received 25,000 applicants for a standard engineering position, and its staffing people declared not one qualified. Cappelli asks the obvious question: could that really be possible? The pattern here is the same one visible in Ben Venue Laboratories, which found only 47 of 3,600 applicants qualified for 100 Ohio openings—though there the failure was basic reading and math tests, a genuinely different case. The engineering example is the pure inflated-requirement case: with an abundance of applicants, the employer demands someone who fits perfectly and can perform from day one without ramp-up, so the screen rejects everyone. The tell is the ratio itself—when zero of 25,000 pass, the filter is broken, not the pool. Recruiters who wrote to Cappelli confirmed their own organizations had 'shortages' precisely because standards were unrealistic and they would not train candidates who could otherwise do the work.

• If a competent current employee (or you) would fail your own screen, then the requirements are inflated—cut them.
• If a requirement names a specific prior title or an exact prior machine/tool rather than a capability, then delete it and describe the capability instead.
• If you have thousands of applicants and 'none qualified,' then treat it as a filter defect first, not a supply problem.
• If the real objection is candidates won't work at your wage, then label it a pricing decision, not a skills shortage.
• If labor is abundant, then deliberately hold requirements steady instead of letting the buyer's market inflate them.

• Treating hiring like the Home Depot model—assuming the job is a machine part with precise specs a candidate must fit perfectly. Tell: language like 'perfect fit' and refusal to consider that the same tasks can be done many ways.
• Letting keyword-based applicant-tracking software define requirements by default. Tell: qualified people rejected for missing an exact phrase or a company-unique job title.
• Blaming schools and the education system for a shortage of experienced workers. Tell: CEOs citing bad education while simultaneously refusing to hire recent graduates or train anyone.
• Waiting for the perfect candidate because abundance is dazzling—the prom-date problem. Tell: standing vacancies left open for months while many merely-qualified applicants are passed over.
• Confusing 'can't find' with 'won't pay market rate.' Tell: complaints about no candidates alongside below-market wages, framed as a skills gap rather than a price.

• ▸Requirement as written — copy each listed requirement verbatim
• ▸Capability or credential? — mark whether it names a real capability or just a prior title/tool
• ▸Can the job be done without it? — yes/no; if yes, flag for removal
• ▸Keyword-screen dependency — does the ATS auto-reject on this exact term? yes/no
• ▸Self-test result — would a competent insider pass this requirement? yes/no
• ▸Market condition adjustment — is this bar inflated by current applicant abundance? yes/no
• ▸Wage vs. skill check — is the real barrier pay rather than qualification? yes/no
• ▸Make-or-buy decision — will you train for this or require it pre-loaded?
• ▸Revised requirement — rewrite as a capability at the true minimum needed

1. 1Pull the actual market wage before diagnosing any hiring problem: look up the occupation in Bureau of Labor Statistics data and record the average wage for that job title and region, not the wage you currently pay.
2. 2Compare your posted wage against that BLS figure and compute the gap as a percentage — a 30-50% shortfall is the same order of magnitude that stalled Mechanical Devices at $13/hr against a $19+ average.
3. 3Add every non-wage cost the offer imposes on applicants (relocation, unpaid ramp-up, shift structure) and treat those as effective pay cuts — the Emirates job paid $2,500/month but required moving to Dubai.
4. 4Re-read your own 'talent shortage' complaint as a price statement: convert 'we can't find anyone' into 'we can't or won't pay the market price,' the way you would with any other scarce good you can buy at market rate.
5. 5Ask the diagnostic question directly of the hiring manager: 'Have we tried raising wages?' If the answer is 'our wages are already high enough' without a BLS comparison, that is the tell of a below-market offer, not a skills gap.
6. 6Run the 'make or buy' calculation: price out hiring at market wage versus hiring basic-ability candidates and training them (apprenticeship, ramp-up time) and choose the cheaper path instead of holding the vacancy open.
7. 7Track the cost of the unfilled vacancy — Mechanical Devices' 40 open machinist jobs were choking sales by ~20% — so the true cost of waiting becomes visible and stops being invisible.

Mechanical Devices, a parts supply company, reported it could not fill 40 machinist jobs, and the deficit was holding back sales by roughly 20%. The story circulated as proof of a skills gap. The jobs paid $13 per hour, which sounds respectable in isolation. But the Bureau of Labor Statistics average wage for that machinist work was more than $19 per hour — almost 50% higher than what the company offered. Once you set the offer against the market rate, the 'shortage' dissolves: the applicant pool shrinks because the price is below market, exactly as it would for any good priced under its going rate. The company was shopping around to see who would take the job below market, which is a pricing decision, not evidence of a skills problem in the workforce.

• If your posted wage sits meaningfully below the BLS average for the occupation, then the shrunken applicant pool is a price problem — stop calling it a skills gap.
• If the hiring manager says 'our wages are high enough' but cannot cite a market comparison, then treat that as an unexamined assumption and force the BLS lookup before approving another job posting.
• If the job imposes large non-wage costs (relocation like Dubai, unpaid time), then discount the headline pay by those costs and re-ask whether it clears the market — few applicants, like the ~50 who showed for Emirates fairs, signal the discounted offer fell short.
• If training a basic-ability hire costs less than paying the market wage for a finished candidate, then buy basic skills and build them internally rather than leaving the seat empty.
• If you would not accept the offer yourself for that move or that work, then don't attribute the empty seat to applicants who 'just don't want to work.'

• Reading vacancies as proof that candidates are deficient — 'the jobs would have been filled otherwise.' Pundits have said this after every recession and are always wrong; the tell is a shortage claim with no wage data attached.
• Comparing your wage only to itself over time ('we pay more than we used to') instead of to the current market. The tell is wages that 'have not gone up in years' defended as adequate.
• Ignoring embedded costs that applicants must absorb, so a $2,500/month Dubai job looks like laziness rather than a below-market net offer. The tell is blaming low turnout at job fairs while dismissing the relocation hurdle.
• Under-reporting the wage problem in surveys — only about 11% of employers in the Manpower survey admitted applicants wouldn't take offers at their wage, but the real share is 'much, much higher' because people rarely name their own fault. The tell is a survey answer of 'talent shortage' from an employer who never tested a higher wage.
• Confusing 'expensive' with 'scarce': you can buy all the diamonds or software engineers you want at the market price. The tell is treating a high market wage as an unavailable-supply story rather than a price you refuse to pay.

• ▸Job title as posted — write the exact title used in the vacancy.
• ▸Your current offered wage — state hourly or monthly rate before benefits.
• ▸BLS market average for this occupation/region — enter the looked-up figure.
• ▸Wage gap % — (market average − your offer) ÷ market average, expressed as a percent.
• ▸Non-wage costs on the applicant — list relocation, unpaid ramp-up, shift burdens.
• ▸Effective offer after costs — subtract non-wage costs and re-compare to market.
• ▸Cost of leaving this seat empty — estimate lost sales/output per week open.
• ▸Make-or-buy cost — compare market-wage hire vs. train-a-basic-hire cost.
• ▸Manager's stated reason for non-fill — record it, then flag whether it cites market data.

1. 1Map the exact language of the job posting before you write anything: pull every noun-phrase skill, tool name, certification, and job title from the description, because the software scans for those literal key words and rejects applications missing them.
2. 2Mirror the vacancy's job title inside your résumé even when it's a title 'unique to that particular company'—the reader who was told he was 'perfect' still got rejected because his prior title didn't match the vacancy's.
3. 3Strip formatting that machines can't parse: don't use headers or footers, since these jam the parser and can cause the software to drop or misread your content.
4. 4Answer the wage question defensively. When the system asks the wage you'll accept, understand you're in an auction—guess too high and the software kicks you out, guess too low and you lock in below-market pay. Research the market rate and answer at it rather than bidding against yourself.
5. 5Treat every questionnaire item as a pass/fail hurdle, not a preference. A typical application has dozens of questions each 'necessary' to be considered fully qualified; even 14 requirements is enough to reject 25,000 applicants for a standard engineering role, so account for the multiplicative screen-out.
6. 6Assume no human sees a rejected application. Because software has replaced most recruiters and applications are done online, you 'rarely talk to anyone,' so there is no chance to 'iron out these little misunderstandings over a cup of coffee'—the résumé must clear the screen unaided.
7. 7Include even trivial tool experience by exact name. Candidates get rejected for lacking 'two years of experience' on a report-formatting tool a decent worker masters in under an hour, so list the specific tools named in the posting if you've touched them at all.

A company posts a 'reasonably standard engineering position' and receives 25,000 applicants, yet staffing people report that not one is qualified. Cappelli tests whether this is possible: it is. Model the application as a series of pass/fail questions, each 'necessary for someone to be considered fully qualified.' With just 14 such requirements—fewer if some hurdles are highly specific, like a title only that company uses or a niche database tool—the multiplied odds of any single applicant clearing every hurdle collapse toward zero. A separate Philadelphia-area HR executive applied anonymously to a job at his own company and 'didn't make it through the screening process.' The takeaway: the shortage is manufactured by stacked hurdles and capricious algorithms, not by an absence of capable people.

• If the posting's job title is company-specific, then reproduce that exact title on your résumé rather than your equivalent prior title—an 'accurate' title that doesn't match still gets rejected.
• If a required tool or skill is named that you can learn in hours, then still list it by name if you've used it at all, because software counts absence of the key word as disqualification.
• If the system asks your acceptable wage, then answer at researched market rate—too high triggers automatic rejection, too low saddles you with below-market pay.
• If you're an employer and no candidate clears the screens, don't manually pull 'almost' candidates back in—doing so treats applicants inconsistently and raises adverse-impact legal risk under antidiscrimination rules.
• If a job description is vague, expect a flood of applications and heavier automated screening downstream; write tighter, defensible qualifying requirements instead of relying on the software to sort the largess.

• Employers demand candidates who 'can step right into the job' with no training, so they require you to already hold the job to get the job—the tell is 'prior success operating similar cotton candy machines' for work that is not rocket science.
• Stacking dozens of individually reasonable questions into hurdles that no real person can clear—the tell is a huge applicant count (25,000) paired with 'not one was qualified.'
• Letting anonymous software decide wages via an auction, then labeling underbidders who decline as 'unqualified'—the tell is calling it an applicant shortage when it's really an unwillingness to pay the market price.
• Vague job descriptions written because hiring managers can't agree on what they want—the tell is a posting that invites everyone to apply and then clogs the system with its own largess.
• Assuming a human will rescue a near-miss application—the tell is believing an HR manager who says you 'have what it takes' will override the software; they won't, and the position stays open.

• ▸Exact vacancy title (copy verbatim from posting; use on résumé even if company-specific)
• ▸Key-word list (every named skill, tool, certification pulled from the description)
• ▸Missing key words (items you can honestly claim but omitted—add them)
• ▸Tool/experience gaps (named tools or year-requirements to address or reframe)
• ▸Formatting scrub (confirm: no headers, no footers, machine-parseable text)
• ▸Qualifying requirements met (count of listed requirements you clear vs. total)
• ▸Wage answer (researched market rate to enter—not high, not underbid)
• ▸Human contact available? (yes/no; if no, résumé must clear screen unaided)
• ▸Adverse-impact note for employers (are these hurdles defensible or just picky?)

1. 1Frame the vacancy as a make-or-buy decision explicitly: list which skills you can realistically buy on the outside market at a wage you can afford, and which you must build internally through training or ramp-up.
2. 2Separate what the job truly requires from what you want ready-made. Distinguish job-specific and work-based skills (only gained by doing the job) from academic credentials, because you cannot buy work experience off the shelf.
3. 3Cost the two paths side by side: the fully loaded cost of hunting for a ready-made hire (months of vacancy, higher wages, elite screening) versus hiring for basic ability and adding training or ramp-up time.
4. 4Quantify your current training investment in hours and weeks per worker per year, then compare it to what the job now demands — especially where automation has raised skill requirements above the old hands-on baseline.
5. 5Address the flight risk directly rather than using it as a reason not to train: structure the financing so the investment pays off even if some trained workers leave (shared cost, apprenticeship arrangements, or partial payback structures).
6. 6Study high-apprenticeship systems (Germany, Switzerland, Scandinavia) as models: they run tighter labor markets yet report fewer skill-gap complaints because they teach work-based skills through structured apprenticeships.
7. 7Audit your screening so it does not defeat the training path: stop rejecting candidates who lack an exact prior job title or key word when they have the basic abilities you could train up.

Mark Collins, a manufacturing expert cited in the book, shows the mechanics. A generation ago virtually all major manufacturers ran large-scale apprentice programs; now, with about 12 million manufacturing jobs still in the US, only 18,000 apprentices remain — just over one-tenth of 1 percent of that workforce. Over the same period, computer-driven machining raised the skills those jobs require, adding IT demands on top of hands-on machining. Yet Collins observes that supplier-side investment in training new workers fell over 25 years, effectively inversely proportional to the investment in automation. The result: less preventative maintenance, more emergency breakdowns, and lower-skilled workers. The right move was to raise training investment as automation raised requirements; instead, employers cut it and then complained of a skills gap that was really a training gap.

• If a skill is unique to your company or cannot be found in the outside market, then build it through training or ramp-up rather than expecting to buy it ready-made.
• If candidates with the exact skills are plentiful and their wages are falling, then buying ready-made makes sense; if they are scarce and expensive, then hiring for basic ability and training is the cheaper path.
• If you raised automation or complexity in a role, then raise training investment proportionally — otherwise expect breakdowns and lower-skilled workers.
• If your stated reason for not training is that workers might leave, then redesign the financing (shared cost or apprenticeship) instead of abandoning training, because in a market where everyone wants trained workers and no one pays to train, no one gets them.
• If you say 'we can't find anyone,' test whether you actually mean 'we won't pay the market wage' — a shortage does not exist at a price you refuse to pay.

• Demanding candidates fill a role instantly with zero ramp-up — the Catch-22 where you must already have the job to get the job. Tell: rejecting able applicants because they lack current work experience.
• Blaming the education system while the real gap is your own missing training. Tell: CEOs citing bad schools, though hiring managers complain about lack of work experience, not academic skills.
• Treating employee flight as a blanket excuse not to invest. Tell: nearly a fifth of employers in the Business Roundtable survey cite risk of workers leaving, while 76 percent cite cost.
• Letting applicant-tracking software screen out qualified people on key words or exact prior job titles. Tell: 'perfect for the position except my old title didn't match the company's unique title.'
• Confusing safety and vendor-equipment orientation with real job-specific training. Tell: your training log is dominated by workplace safety and 'here's how to work this new copier' rather than skills that build capability.

• ▸Role and its core skills: list the specific work-based skills the job requires
• ▸Buyable on market?: for each skill, yes/no and the wage needed to buy it ready-made
• ▸Unique to us?: mark skills so specialized they cannot be found outside
• ▸Cost to buy: months of expected vacancy + wage premium + screening cost
• ▸Cost to build: training hours/weeks needed + ramp-up time to productivity
• ▸Automation delta: has complexity risen since last staffing? raise training accordingly
• ▸Flight-risk financing: how the training cost is shared or recouped if worker leaves
• ▸Screening check: does our software reject able candidates on title or key words? fix