Friday morning, the Head of Talent at a 500-person SaaS company opens the req queue. Twelve open roles. The customer support specialist posted Tuesday is already at 380 applicants. The recruiter assigned spent most of yesterday triaging the inbox and still hasn't surfaced a shortlist. Two other reqs haven't been touched since Wednesday. In ninety minutes the hiring manager will ask why nothing is moving on customer support, and the honest answer is that the team got buried.

The instinct, when a role hits 380 applicants, is to staff up the screening or buy software that promises to read every resume in under a second. Neither instinct is wrong, but neither addresses what's actually broken. The volume is the symptom. The structural failure is the screening process itself, designed for slates of 30 to 50, now absorbing slates of 300 to 500.

The volume problem isn't a volume problem

Applications per hire have climbed sharply since 2021, with many high-volume roles now drawing several hundred applicants per opening. The recruiters absorbing that volume haven't grown in proportion. SHRM's 2025 Recruiting Benchmarking report finds over half of organizations have recruiters managing about 20 requisitions each, with larger firms running heavier loads.

Pinch that math. Twenty requisitions per recruiter, two hundred applicants per req on the lighter end, and the work sits somewhere around 4,000 applicant evaluations per recruiter at any given moment. Even at a generous 90 seconds per resume, that is a hundred-hour task. It is not a workload that can be solved by working harder, and it is not a workload that anyone should be asked to solve alone. The system needs to be different. Not the people in it.

Define what "qualified" means before applications land

The biggest gain comes upstream of the resume pile. Most hiring teams write a job description, post it, and then ask the recruiter to figure out what a yes looks like on the fly. By the time application 200 lands, the criteria have drifted, the hiring manager has added a new "nice to have," and the same role gets evaluated differently on Tuesday than it did on Friday.

For a high-volume role, before the posting goes live, the recruiter and hiring manager need to agree on three things, in writing:

  • The three to five criteria that make someone a yes at the screening stage. Not the dream profile. The screening floor.
  • The disqualifiers an automated step can apply without human judgment (work authorization, certifications required by law, geographic constraints).
  • The two or three questions whose answers actually predict whether a phone screen will go anywhere. These should be answerable in writing or in a structured screening conversation, not invented mid-call.

This is the work the system can amplify. When the criteria are vague, more automation just produces confidently wrong shortlists faster.

What to automate, what to keep human

Automate the structured layer. Resume parsing against the screening floor. Disqualifier checks. The first structured screening conversation, where every candidate answers the same questions against the same criteria the hiring manager already signed off on. Scheduling the follow-up. Sending status updates so candidates aren't waiting in silence.

Keep the judgment layer human. The yes-or-no on borderline candidates. The conversation with the hiring manager about how the slate is shaping up. The closing call with the candidate you actually want. The decision on whether to widen the criteria when the funnel runs dry, or to push harder on the candidates already in flight.

This split matters because the structured work is where inconsistency creeps in. Two recruiters reading the same resume on a Friday afternoon make different calls. The same recruiter reading the same resume Monday morning versus Friday evening makes different calls. A structured screening conversation that asks every candidate the same questions, scored against the same criteria, doesn't have that drift. That is what an AI screening agent can do well, and it is where humans, working at volume, run out of consistency.

What it cannot do is decide who to hire. Sia, the Eximius screening agent, surfaces qualified candidates against the criteria the recruiter set. The recruiter still runs the shortlist meeting. The hiring manager still owns the offer. The system supplies cleaner inputs to those decisions, faster.

Track the funnel, not the activity

At 500 applicants, the trap is to start measuring how many resumes the team processed. That measure rewards motion, not outcomes. The metrics that tell you whether the process is working are funnel metrics: what percentage of applicants pass the screening floor, what percentage of screened candidates the hiring manager wants to interview, how many of those reach offer, how many offers convert.

For a baseline, Ashby's analysis of more than 54 million applications puts the recruiter-screen passthrough rate around 35% overall. If your floor is significantly higher than that on a high-volume role, the criteria may be tighter than the market supports, and you're filtering out candidates the hiring manager would have wanted to talk to. If it is much lower, the screening criteria are doing less work than the resume pile would suggest, and the burden is moving downstream to the hiring manager.

Either way, the funnel tells you where to fix. The activity count tells you nothing except that people were busy.

The recruiter's day, when the process works

When screening is structured and the right things are automated, the recruiter's calendar shifts. Less time triaging inboxes. Less time answering "what's the status" emails. More time on the conversations that move offers across the line: aligning with hiring managers, debriefing panels, closing finalists. The work that actually requires a recruiter is the work that gets done.

That is the test of any screening system, including this one. Not whether it processed 500 applicants. Whether it gave the recruiter back the hours the hiring manager and the finalist actually need from them.

Want to see what structured screening looks like on a 500-applicant role? Book a pilot and we'll run your next high-volume req through the Eximius workflow.