Why Job Boards Haven't Changed Since 1994
Monster launched in 1994. Everything since has been a UI update on the same broken model. AI just made the core failure catastrophic.
Monster launched in 1994. That’s not a throwaway stat — it’s the year the modern job board model was set, and it hasn’t fundamentally changed since.
The format: employer pays to post a job, candidates apply for free, platform collects applications and delivers them to the employer. Volume is the product. More applicants means more activity, more data, more reasons to justify the price of the posting.
LinkedIn Jobs arrived. Indeed scaled to dominate. Glassdoor added reviews. ZipRecruiter brought programmatic distribution. The UI improved. Matching algorithms got more sophisticated. But the underlying structure stayed the same: employer pays, pile forms, employer deals with the pile.
What the AI era did to this model
AI didn’t fix the pile. It made it worse.
When applying to a job takes 10 minutes instead of an hour — when cover letters write themselves, when your resume can be reformatted to match any job spec — the volume problem becomes catastrophic. A single posting that used to attract 200 applications now gets 2,000. Most of them are AI-generated. Most of them look qualified on paper. Almost none of them have been meaningfully considered by the person who sent them.
The signal-to-noise ratio collapsed.
Employers responded by buying better filtering tools, stricter ATS gatekeeping, and resume screening software designed to thin the pile before a human sees it. The candidates who adapted sent more applications, to more jobs, with better AI-generated copy. The pile got bigger. The tools got stricter. The system got worse for everyone.
Who the current model works for
Nobody, if we’re honest.
Employers pay to generate a pile they then need to separately pay to manage. The best candidates are hired by the companies with the fastest screening processes — which are usually the best-resourced companies, not the ones who most need great people.
Candidates spend real time on applications and hear nothing back. Not because employers are indifferent, but because there is genuinely no infrastructure to respond to 400 people with care. Silence is the rational outcome of an irrational volume.
The job board keeps the money either way.
What a different model looks like
The core fix isn’t better matching or smarter AI filtering. It’s structural.
If you cap the number of applications a role can receive, the pile doesn’t form. If screening is built into the posting process — not sold as a separate product — employers don’t have to build their own infrastructure. If every applicant is guaranteed a response, the candidate experience changes fundamentally.
None of this is technically difficult. It’s a design choice. The platforms built before AI had no reason to make it. The ones built after have a chance to.
That’s what we built ORecruiter to be. A job board — with different structural decisions.