Job postings are pulling in far more applications than they did a year ago, and most of the increase has nothing to do with more people looking for work. AI writing tools let one candidate apply to hundreds of jobs in the time it used to take to apply to ten, and recruiters are the ones absorbing the flood. For staffing agencies still matching candidates to positions by hand, that math gets ugly fast.
The numbers behind the flood
LinkedIn now processes roughly 11,000 job applications every minute — a 45% jump in a single year, driven largely by generative AI tools that write and submit applications automatically, according to LinkedIn data reported by The New York Times and eWeek. A single remote job posting can pull in more than 1,200 applications within days, and most of them sit unread, Ars Technica has reported.
Recruiters are feeling the size of that pile directly. Ashby's May 2026 analysis of more than 100 million applications and 200,000 jobs found that applications per hire have tripled since 2021, with the average open role now pulling in more than 300 applications. The candidates didn't get more numerous. They got more productive.
More resumes, not better matches
The extra volume hasn't made hiring easier — by most accounts, it's made it slower. A Robert Half survey released this year found that 67% of HR leaders say reviewing AI-generated applications has slowed down their hiring process, and about one in five report delays of two weeks or more because of it. Nearly two-thirds also say it's gotten harder to tell whether a candidate's listed skills are actually real.
That's the uncomfortable part. A resume can be well-written and still not tell you much. When every applicant has access to the same tools that produce a clean, professional-sounding document in minutes, a polished resume stops being a signal of anything except that someone used an AI tool.
What this looks like for a staffing agency
A single company hiring for one open role feels this as a slow week. A staffing agency feels it differently, because agencies are usually filling several positions across several clients at the same time, each with its own shift, certification, or location requirement, pulled from a jobseeker pool that keeps growing every week.
Doing that matching by scanning a shared inbox or a spreadsheet was already a slow, manual process before application volume tripled. Add a pile of AI-polished profiles that all look similar on the surface, and a recruiter can spend an entire afternoon just figuring out which five people in a stack of two hundred are actually worth a phone call.
Keyword matching was built for a different pool
Most of the tools agencies lean on for a first pass — a search box, a filter on a spreadsheet column, a resume database search — work by matching keywords. That approach made sense when a resume's wording was a rough proxy for a candidate's actual background. It makes a lot less sense now that AI writing tools are specifically good at guessing which words an applicant tracking system is scanning for and working them in.
The result is a pool where keyword density and actual fit have drifted apart. A candidate who's missing a required certification can still write a resume that hits every keyword in the posting. A candidate who's exactly right for a role can miss the keywords entirely just because they described their experience differently. Neither of those is a problem a bigger, faster keyword search solves.
| Keyword search | Semantic matching | |
|---|---|---|
| What it checks | Whether specific words appear in the text | What the profile actually describes |
| Easy to game with AI writing tools | Yes | Harder |
| Misses candidates who phrase things differently | Often | Less often |
| Gets slower as the pool grows | Yes, linearly | Not by design |
What holds up better against a pool like this is matching that looks at what a candidate's profile actually describes — their experience, qualifications, and history — rather than whether a specific phrase from the job posting shows up somewhere in the document. That's a different kind of search entirely. Instead of asking "does this word appear," it asks "how close is this candidate's actual background to what this position needs," which doesn't get fooled by a resume that's been optimized to include the right vocabulary without the underlying substance.
For a recruiter working through a stack that's tripled in size since last year, that difference isn't academic. It's the gap between spending an afternoon reading two hundred resumes and spending ten minutes reviewing a shortlist that was already ranked by fit.
Rhythence's Placements tool works this way: it matches jobseekers to open positions using semantic similarity against verified profiles, not keyword search, so a match is based on what a candidate's background actually is rather than how closely their resume happens to echo the posting. If your agency is feeling the size of the applicant pool this year, book a demo to see how that matching works against your own jobseeker data.
See Placements in action
Get the right person in the right role. Discover how Rhythence helps staffing agencies eliminate admin tax and streamline operations.
Get staffing agency insights delivered
Join operational leaders and founders receiving practical breakdowns on placements, compliance, and agency scaling.
Frequently Asked Questions
Why are job postings suddenly getting so many more applications?
AI writing tools make it fast and cheap to apply to many jobs at once, so the same job seekers are sending out far more applications than before. It's not that more people are job hunting — it's that each of them can apply to more roles in less time.
Does keyword screening still work against AI-written resumes?
Less reliably than it used to. AI writing tools are good at anticipating which keywords an applicant tracking system is scanning for, so a keyword match tells you less about actual fit than it did a couple of years ago.