TL;DR: Applicants describe a 40% ghosting rate and a 5% response rate, and spend their week on listings that were never going to produce an interview. The dominant cost is not rejection — it is applying into postings that no longer have a role behind them. Nothing in the current stack tells a candidate which listings are worth the hour.
The evidence
PainHunt's HR tech, recruiting and hiring categories hold 806 high-scoring signals (10+/15), average score 11.4, average pain intensity 7.3/10 across 4,333 extracted pain points. Sources are Mastodon (190), Medium (159), Reddit (139), BlueSky (101), Hacker News (51) and RemoteOK (42) — a mix of practitioners and people currently job hunting, rather than app reviews.
The candidate-side cluster inside it is unusually specific about numbers:
- A 40% ghosting rate and a 5% response rate. Not "some employers are slow" — a majority-silent process, described as prolonging unemployment.
- High volume of spam and fake postings making legitimate opportunities hard to identify, and time-consuming to sort.
- Application volume for junior roles is unmanageable on the employer side too. The same flood that makes candidates invisible makes recruiters unable to find genuine talent without invasive filtering.
- Monitoring tooling made the experience worse, not better. Eye-movement and browser-focus tracking during interviews is described as hostile — the industry's answer to volume has been surveillance, aimed at the wrong end of the problem.
The second and fourth points matter together. Both sides are drowning, and every tool built so far has been sold to the employer.
Why now
Posting became free and applying became one click. The cost of listing a role fell to roughly zero, and so did the cost of applying. Both volumes exploded, and the signal-to-noise ratio on both sides collapsed with them.
AI removed the last friction from applications. A tailored cover letter now takes seconds. Employers respond by filtering harder, candidates respond by applying wider, and the loop tightens with nobody choosing it.
Pay transparency laws created a checkable fact. In jurisdictions requiring salary ranges, a posting that omits one is either non-compliant or not really located where it claims. That is the first externally verifiable signal this market has had, and it is not being used.
The wedge
The buildable thing is a freshness and legitimacy score on the listing, sold to the applicant.
- Score what is checkable, not what is claimed. How long has this ad run; how many times has it been reposted verbatim; does the same req appear on the company's own careers page; is there a named recruiter who exists elsewhere; is a salary range present where law requires one. Each is weak; the combination sorts a list.
- Rank the list, don't block it. "Apply here first" is a useful product. "This job is fake" is a defamation risk and will be wrong often enough to destroy trust. The output should be an ordering, with the reasons shown.
- Track outcomes and close the loop. Ask users what happened — reply, interview, silence — and feed it back per company. After enough volume this becomes the only dataset that measures employer responsiveness from the applicant's side, and that dataset is the actual moat. Nothing else here is hard to copy.
- Publish employer response rates. The consumer-facing version of the outcome data. It creates the reputational pressure that changes behaviour, and it is the reason the product gets talked about — but it is also what makes job boards unwilling to partner, so decide early whether that trade is worth it.
The entry point is a browser extension over the boards people already use, not another job board. Competing on listing volume against incumbents is the losing version of this.
Risks and honest caveats
Job seekers are a famously bad market to sell to. They are cost-sensitive, and by design they churn out the moment the product works. Lifetime value is bounded by the length of a job search. Any business here is either very cheap and very high volume, or it monetises the outcome data rather than the user.
Scoring is inference, and inference will be wrong. Calling a real posting stale sends someone past an opportunity they should have taken. The product has to be honest that it ranks rather than judges, and calibrated conservatively — which makes it feel less decisive than a competitor willing to overclaim.
Data access is adversarial. The signals need scraping boards whose terms discourage it, and boards can change markup or add blocking at any time. This is a real and permanent operating cost, not a launch problem. The adjacent PainHunt cluster on verifying contractors before you hire them has the same shape without the scraping dependency.
The behaviour may be rational for employers. Evergreen pipelines and compliance postings serve real purposes. A tool that only frames them as deception will be dismissed by the people who could most easily fix the problem.
Where this came from
This is one cluster inside PainHunt's HR and recruiting categories. The Pain Point Browser shows the underlying signals with their intensity and commercial scoring, and the Idea Validator will score a specific version of this idea against the same dataset. The employer-side view of the same collapse is how to interview when AI can generate a perfect portfolio — worth reading alongside, because the two describe one problem from opposite ends.