Tool
Job-Search Claim Checker
Paste a job-search statistic and find out whether it has been traced to a source — and what that source actually measured.
Free, no account required. Runs entirely in your browser — nothing you type is transmitted or stored.
Matching runs entirely in your browser. Nothing you type is sent anywhere, stored, or logged.
Type a claim above to search 12 evidence reviews.
Every claim in the index (12)
- 75% of resumes are rejected by an ATS before a human ever sees them.Unverified · Applicant tracking systems
- More than 90% of employers use software to filter or rank candidates before a human reviews them.Supported · Applicant tracking systems
- Applicant tracking systems automatically screen out resumes that show an employment gap.Supported · Applicant tracking systems
- Most resumes never reach a human being.Context dependent · Applicant tracking systems
- Recruiters spend only six seconds looking at a resume.Partially supported · Resumes
- Including a cover letter increases your chance of getting an interview.Partially supported · Resumes
- 80% of jobs are found through networking.Unverified · Networking and referrals
- 70–85% of jobs are never publicly advertised.Unverified · Networking and referrals
- Employee referrals make a candidate several times more likely to be hired.Context dependent · Networking and referrals
- Identical resumes get different responses depending on the applicant's name.Supported · Applications
- Only 2% of applicants are invited to an interview.Context dependent · Applications
- It takes 100 to 200 applications to get a job.Unverified · Applications
What this does, and what it will not do
This searches the evidence reviews published on this site. It matches on the wording of the claim, on the normalised version of it, and on the variants we have recorded, so a paraphrase usually still finds the right review.
If nothing matches, it says nothing matches. It does not generate an assessment, estimate a verdict, or offer a plausible-sounding answer assembled from adjacent claims. An invented verdict about an unreviewed statistic would be the same failure this library exists to document, delivered faster.
There is no language model behind it and no request leaves your browser. The index is shipped with the page, which at this size is smaller than a round trip to a server would be.
If a claim is not here
The library is deliberately small. Reviews are published when a claim has been traced as far toward its original source as we can get, and not before; a great many researched claims remain unpublished because the trail is still open.
The absence of a review is not evidence about the claim. It means we have not finished checking.