Recruiters want job applications to get harder again

Three years ago Andrew Stockwell, then head of people at software-buying company Vendr, could post a job listing and expect to review a few dozen applications, up to 100 on a good run. In the last year, he says, that changed: within a day or two of posting, hundreds of applications arrived, sometimes topping a thousand. A good chunk were outright fake, and many more looked AI-written, stretching the truth and making candidates hard to tell apart. Stockwell says he ended up paying his talent-acquisition staff simply to sort through the pile.

Ophir Samson, head of voice AI at Greenhouse (a recruiting platform he joined this year when it acquired his AI-interviewing startup), says the industry did this to itself. A year ago, he says, every recruiter wanted applying to be as frictionless as possible. What they got instead was 2,000 applicants in 24 hours for a single job. Now, he says, recruiters are telling him the opposite: they want friction back, deliberately, to cut the flood down to a manageable size.

The numbers back up the flood. LinkedIn told WIRED that submissions per applicant on its platform are up 46 percent compared with February 2020, and up 22 percent since the launch of ChatGPT in late 2022. Samson says recruiters routinely describe getting 1,000 applications for a role while knowing only about 30 of them are serious. Meanwhile the supply of jobs has shrunk: US job openings peaked at a record 12.3 million in March 2022, per Bureau of Labor Statistics data, and have hovered around 7 million (give or take 500,000) since mid-2024. Jane Curran, chief transformation officer at real estate firm JLL, contrasts the two eras: during the pandemic hiring boom, “everyone was job hopping, because you literally could have three offers in an afternoon,” she says; now “it is the polar opposite.”

Tools that make applying easier keep multiplying. Companies including JobAssist, Sonara, and Ladder's Apply4Me market themselves on submitting, in their words, “10x as many applications with less effort than one manual application.” LinkedIn's own Easy Apply and one-click features pushed in the same direction for years. LinkedIn says it has responded by capping automated and low-quality applications, and this month is rolling out a feature that tells applicants who appear underqualified for a role that they probably are not a good fit, and points them to other openings.

Recruiters are fighting the volume with AI of their own. Applicant tracking systems have screened resumes automatically for at least a decade, and Samson says many companies now use AI agents to run first-round interviews, sorting the roughly 30 serious candidates in a 1,000-application pile from the rest. Curran expects some companies to add strict knock-out questions or move skills testing earlier in the process. But Tessa White, a former HR executive who left corporate America in 2018 and now advises job seekers to her 800,000 TikTok followers, is not convinced any of this fixes the underlying problem. “Every time we seem to strive for efficiency, we seem to give up quality,” she says, arguing that AI and applicant tracking systems, however high-tech they sound, remain an outdated way to judge people's actual potential.

The flood has pushed recruiters back toward referrals and internal hires as the more trustworthy channel, a shift Stockwell says favors candidates whose backgrounds already resemble a company's existing staff, at the expense of workforce diversity. Stockwell is now living the mess from the other side: he left Vendr in February and is job hunting himself, treating the search as its own job, networking in person because he estimates that applying online leads to a phone interview less than 2 percent of the time.

Key facts

LinkedIn says submissions per applicant are up 46 percent versus February 2020 and up 22 percent since ChatGPT's late-2022 launch.
Recruiters describe receiving around 1,000 applications per role while judging only about 30 of them serious; Greenhouse's Ophir Samson says one employer got 2,000 applicants in 24 hours after trying to make applying frictionless.
US job openings peaked at 12.3 million in March 2022 (Bureau of Labor Statistics) and have run around 7 million, give or take 500,000, since mid-2024.
Recruiters are now asking Greenhouse for deliberate friction; the company and others are pushing AI interview agents and stricter screening to cut the applicant pool down to serious candidates.
LinkedIn is rolling out a feature this month that tells apparently underqualified applicants they probably are not a fit and suggests other jobs, on top of existing limits on automated, low-quality applications.

Why it matters

The article documents a feedback loop specific to the AI era: tools that write tailored resumes and autofill applications in seconds, plus platforms that promise a one-click apply, have collapsed the cost of submitting a job application toward zero. Recruiters spent years optimizing for exactly that outcome, and now describe the result, application volumes several times what they were three years ago, as unworkable. The response from parts of the recruiting industry is not more automation to cope with volume, but a deliberate push to reintroduce friction, which is a notable reversal of a decade-long trend toward frictionless hiring funnels.

Who it affects

Recruiters and talent-acquisition staff, who now spend their time triaging thousands of submissions instead of sourcing candidates. Job seekers, who face lower odds per application (Stockwell estimates under 2 percent of his own online applications reach a phone interview) and a market where referrals and internal hires increasingly beat cold applications. Employers broadly, caught between complaints that they can't find good people and applicants' complaints that they can't find jobs, per Tessa White. AI vendors on both sides of the transaction: resume and autofill tools like JobAssist, Sonara, and Apply4Me driving volume up, and recruiting platforms like Greenhouse building AI interview agents to filter it back down.

How to use it

For job seekers, the piece is a signal that mass one-click applications carry increasingly low odds and that recruiters are actively re-engineering the funnel to punish that behavior, so networking, referrals, and targeted applications carry more relative weight than they did a few years ago. For recruiters and hiring teams, it names concrete levers already in use elsewhere: knock-out questions, earlier skills testing, and AI-run first-round interviews to separate roughly 30 serious candidates out of a 1,000-application pile. No pricing or specific product terms for any of the named tools are given in the source.

How solid is it

The reporting rests on named, on-the-record sources with specific titles and employers (a recruiting-platform AI lead, a corporate transformation officer, a former HR executive, and a hiring manager turned job seeker), plus two hard data points: LinkedIn's own submission statistics and Bureau of Labor Statistics job-openings figures. That is solid sourcing for a trend story, but the core claim, that recruiters now want friction, is illustrated through a handful of individuals' testimony rather than an industry-wide survey.

Risks and caveats

The article does not quantify how many of the flood of applications are outright fake or AI-written beyond Stockwell's qualitative description ('a good chunk,' 'many more'). It does not say what share of recruiters industry-wide share the friction-is-good view; the sentiment comes from named individuals, not a poll. It does not detail how LinkedIn's new underqualified-applicant feature actually works beyond notifying and redirecting candidates, nor which specific knock-out questions or skills tests companies are expected to adopt. The year Stockwell left Vendr is also not stated, only that it was in February.

“Actually, we kind of want friction. The friction is good. We want to make it harder.”

— Ophir Samson, head of voice AI at Greenhouse