A slow search is a funnel with a clog somewhere

When a role won't fill, the complaint is usually vague: "we're just not getting good candidates." But "not getting good candidates" can mean a dozen different things — too few applicants, plenty of applicants but none worth interviewing, strong interviews that don't convert to offers, or offers that keep getting declined. Each of those is a different problem with a different fix, and you can't tell them apart by feel. The recruiting funnel turns that fog into a diagnosis.

The funnel is just your hiring process drawn as stages, with a count at each one and a conversion rate (also called pass-through rate) between them. It's the natural companion to the broader hiring metrics that matter and to cost-per-hire — but where those measure outcomes, the funnel tells you where in the process the outcome is being decided.

The stages and the rates between them

A typical funnel runs through these stages. Count how many candidates reach each, then divide adjacent stages to get pass-through rates:

  • Applied → Screened (did they clear the initial review / knockout questions?)
  • Screened → Phone screen (worth a recruiter call?)
  • Phone screen → Interview (advanced to the hiring team?)
  • Interview → Offer (the loop said yes?)
  • Offer → Accepted (they said yes back?)

A useful summary number is the overall funnel rate — accepted hires divided by total applicants — but the diagnostic power lives in the stage-to-stage rates, because that's what isolates the bottleneck. There are no universal "correct" conversion numbers; what matters is establishing your own baseline across roles and over time, then watching where a specific search deviates from it.

Reading the funnel: each clog has a different cause

Find the stage where your pass-through rate is unusually low for that role, and the cause usually announces itself:

  • Few applicants at the top. This isn't a conversion problem — it's a sourcing or reach problem. The job isn't getting in front of the right people, or the posting isn't compelling. Revisit your job-board distribution and your job description.
  • Applied → Screened is brutally low. Lots of applicants, almost none worth advancing. Your sourcing channels are bringing volume but not fit, or the job description is attracting the wrong profile. Tighten targeting, not just volume.
  • Screened → Interview leaks badly. Candidates look good on paper but fall apart on the phone. Either your paper screen is too generous, or your phone screen is filtering for the wrong things. Re-align them.
  • Interview → Offer is the chokepoint. You're interviewing plenty but rarely extending offers. Often this is an inconsistent or miscalibrated interview loop, not a candidate-quality problem — the fix is structured interviews and a tighter intake meeting so the team agrees on the bar before they start.
  • Offer → Accepted is low. Your process works right up until the close, then candidates walk. That points at compensation, a slow or clumsy offer, a weak candidate experience, or losing the negotiation.

The discipline is to measure first, then fix. Most teams pour effort into the top of the funnel ("get more applicants!") when the real leak is an offer-acceptance problem at the bottom — and more applicants do nothing for a clog three stages down.

Make the funnel trustworthy

The funnel only helps if the numbers are clean:

  • Define each stage the same way every time. If "screened" means different things to different recruiters, your rates are noise. Consistent stage definitions are what make the comparison meaningful.
  • Segment by role and source. An engineering funnel and a sales funnel behave differently; aggregate them and you'll diagnose the wrong thing. Same with source — the conversion profile of a referral is nothing like a cold job-board applicant.
  • Watch trends, not single searches. One req with a weird funnel might be luck. The same leak across several searches is a process problem worth fixing.
  • Let the system count, not a spreadsheet. Funnel math is only sustainable when stage movement is tracked automatically as candidates progress. When every candidate flows through one pipeline and the reporting computes pass-through rates for you, the funnel becomes a dashboard you actually check instead of a quarterly spreadsheet nobody updates.

Bottom line

A stalled search isn't a mystery — it's a funnel with a clog at a specific stage, and conversion rates tell you exactly which one. Define your stages consistently, baseline your pass-through rates by role and source, and find the stage that's leaking before you spend a dollar fixing the wrong end. The team that diagnoses "we're losing people at the offer stage" and fixes its closing process will out-hire the team that just keeps shouting "post the job in more places." When your pipeline tracks every stage automatically and the analytics surface the rates, the bottleneck stops hiding.