- R-01 Why the best-fit candidate is often never seen Collecting sources 1 sourced, 5 untested
- R-02 Not enough roles, or the wrong roles in the wrong places? Collecting sources 1 sourced, 5 untested
- R-03 How hiring systems and job boards decide who gets seen Open question 0 sourced, 5 untested
- R-04 Why hundreds of applications can still mean no interviews Collecting sources 2 sourced, 5 untested
- R-05 What a good match actually predicts Open question 0 sourced, 4 untested
Contents
Why the best-fit candidate is often never seen
When a role draws hundreds of applications, why is the person best suited to it so often not among those who get a conversation?
A 2021 report from Harvard Business School and Accenture found that 88% of the employers surveyed said qualified, high-skilled candidates are rejected outright because they do not exactly match the hiring criteria.1
Limits: This reflects what employers said in a survey, not a count of candidates who were actually rejected.
The first screen may often be a filter (keywords, job titles, years of experience) rather than a judgement of fit.Untested
People describe the same experience in different words from the job advert, so a good fit can look like a weak one on paper.Untested
Recruiters may have only seconds per application, so the order applications arrive in could matter more than it should.Untested
Referrals and internal moves can be considered before the public pile is read at all.Untested
CV layout can break automated parsing, turning strong experience into missing fields.Untested
Next step
Find primary evidence on how often automated filters, rather than people, make the first cut.
Not enough roles, or the wrong roles in the wrong places?
Is the problem too few openings, or a mismatch between where people want to work and where the openings are?
The 2010 Nobel Prize in Economics (Diamond, Mortensen and Pissarides) recognised analysis of markets with search frictions, which they applied to labour markets to understand unemployment, job openings and pay.1
Limits: This is a general theory of why finding a match is not instant. It does not measure any one job market today.
Some fields attract far more interest than they have openings, while others struggle to fill roles.Untested
Openings cluster by city, seniority and skill, while interest clusters somewhere else.Untested
There are fewer senior roles than senior candidates, so competition rises with experience.Untested
Entry-level roles that ask for experience shut out the people they are meant for.Untested
Many searchers keep to a narrow set of titles even when adjacent titles would suit them.Untested
Next step
Look for official labour-market statistics comparing vacancies and job seekers by field and region in the UK and Canada.
How hiring systems and job boards decide who gets seen
Which choices made by job boards and applicant tracking systems decide whose application is read first, and whose is never read?
Nothing we can stand behind yet.
Ranking by recency or by sponsorship can favour whoever applies first, or whoever pays.Untested
One-click apply makes applying cheap, which raises volume and lowers the attention each application gets.Untested
Knockout questions can reject people automatically before any person looks.Untested
The same role is often posted on several boards, splitting attention and hiding true demand.Untested
Employers and boards want different things: employers want fewer, better applicants, boards want more activity.Untested
Next step
We have not found reliable independent evidence on how ranking works inside the major platforms. Most public claims come from vendors, so we are looking for better sources.
Why hundreds of applications can still mean no interviews
Why do some people send 300 applications and still not get a single interview?
In Greenhouse's 2024 State of Job Hunting report, three in five candidates surveyed (2,500 workers across the US, UK and Germany) said they suspected they had met a ghost job, meaning a listing with no real intent to hire.1
Limits: This measures suspicion, not proof.
The same report said Greenhouse's own analysis classed roughly 18 to 22% of the jobs posted on its platform each quarter as ghost jobs.2
Limits: This covers one vendor's platform in 2024 and uses that vendor's own definition, so it cannot be applied to the whole market.
Volume spreads effort thin: many applications go to roles that are poor fits, which no tailoring can fix.Untested
Some of the roles are not really open: reposted, already promised internally, or kept up to look like growth.Untested
Generic applications compete badly against tailored ones.Untested
Silence is not always rejection: many employers never reply at all.Untested
Targeting fewer, better-matched roles may beat sending more.Untested
Next step
Find independent data on application-to-interview rates and on how many listings are genuinely open.
What a good match actually predicts
Does a high match score predict an interview, or only that a CV resembles a job description?
Nothing we can stand behind yet.
A score can measure overlap in skills and seniority without measuring what a hiring manager values.Untested
A useful score should come with its reasoning so a person can disagree with it.Untested
Fit for the person matters too: interest, location, pay and growth, not only the employer's checklist.Untested
Knowing whether matches lead to interviews needs outcome data, which is sensitive and hard to collect responsibly.Untested
Next step
Work out how any tool could learn which matches lead to interviews without collecting more personal data than it should.