The irrelevant-applicant problem: why one opening pulls thousands of resumes that don’t fit

Post one entry-level opening in India and you don’t get fifty applicants. You get a few thousand, and most of them shouldn’t be there. A customer success role pulls in facility managers, schoolteachers, and people with fourteen years in a field nothing like the one you’re hiring for. The volume gets all the attention. The bigger problem is that a large slice of that volume was never a fit to begin with, and your screening process has no fast way to say so.

Volume is the symptom; mismatch is the disease

Every high-volume hiring post about India opens with the same number: tens of thousands of applications per role, 72-hour application windows, recruiting teams of eight. All true. But framing it purely as a capacity problem hides the part that actually wastes the most time.

Recruiters in India spend an estimated 60–70% of their time filtering unqualified profiles, not evaluating qualified ones. That’s the tell. If the queue were simply large but well-targeted, you’d add reviewers and grind through it. The reason teams burn out isn’t the count. It’s that the count is mostly noise, and a human has to read each line to find that out.

Free and frictionless applications make this structural. When applying costs the candidate nothing, the rational move is to apply everywhere. A ₹30K-CTC opening doesn’t attract 30K people who want that job; it attracts 30K people who want a job. The mismatch is baked into the channel, not the candidate.

Why the resume screen makes it worse, not better

Here’s the uncomfortable part: the tool most teams reach for to handle the flood, resume keyword scoring, is calibrated for exactly the wrong thing.

Most resume scorers reward total years of experience and degree level. So the fourteen-year facility manager applying to your customer success role doesn’t get filtered out. They often score higher than the sharp first-jobber who can actually do the work, because the screen is counting tenure and credentials rather than asking whether any of it is relevant. We wrote more about that scoring failure in why your AI resume scores all cluster at 50–60%. The short version: a relevance-blind screen doesn’t reject the wrong people. It reshuffles them.

That’s how you end up with a “shortlist” that still needs a human to re-read it from scratch. The screen did motion, not work.

What a relevance-first screen actually checks

The fix isn’t a bigger funnel or more reviewers. It’s changing what the first filter measures: from how much a candidate has done to whether it’s relevant to this role.

Three things move the needle, in order:

1. Role-specific knockouts first. Ask the two or three disqualifying questions that define the role before you evaluate anything else. If a job genuinely requires on-site presence in Pune and the candidate can’t relocate, that’s a 60-second conversation, not a 15-minute one. Reject early, reject cheaply, and tell the candidate why.
2. Relevance over volume. Fourteen years in an unrelated field should count for close to nothing against a role that needs none of it. A screen that can’t express “irrelevant” will always inflate the wrong profiles.
3. Communication as the real first signal for entry roles. For most fresher and entry-level hiring, the question that actually predicts success isn’t on the resume at all. It’s whether the person can hold a clear, structured conversation. You only learn that by listening to them, not by parsing a PDF.

None of this requires sorting through more resumes faster. It requires admitting that resume sorting was never the screen.

The math changes when the first filter is honest

In our own pilot campaigns we’ve run first-round screening across a few thousand candidates in a single evening, with the AI voice agent handling a 15–20 minute structured conversation and the recruiter time per candidate dropping by roughly 89% versus manual phone screens. But the throughput number isn’t the point. The point is what gets through. When the first filter measures relevance and communication instead of tenure and keywords, the people who reach a human are the people worth a human’s time.

That’s the difference between processing applicants and screening them. Processing moves a queue. Screening makes a decision, including the decision to say no early, clearly, and to the right people.

If you’re staring at an opening that’s about to pull in a few thousand resumes you don’t have the bandwidth to read, the question isn’t “how do we go faster.” It’s “what is our first filter actually measuring.” Get that wrong and speed just helps you mis-sort faster. Get it right and the flood stops being the problem.

See what a relevance-first first round would do to your own numbers with our screening ROI calculator.