Skip to content
CogniYukti

Analytics

Counting submissions rewards the wrong behaviour.

Any recruiter can hit a submission target. The question is what happened to those submissions, and whether the same thing keeps happening.

Suite C

Ten cards, and one of them is a coaching conversation.

Output, throughput and load — the ordinary measures, per recruiter and against the team. Then the one that is harder to get anywhere else: rejected-candidate pattern analysis, which reads the structured rejection reasons on a recruiter's declined submissions and says what keeps recurring.

That is a coaching card, not a performance score. “Six of your last twenty declines were notice period” is an actionable sentence; “conversion 14%” is not.

One recruiterThemMedian
  • Hours to first submission3122
  • Submitted → shortlisted38%26%
  • Rejected below experience bar197
  • Offers accepted81%79%
They convert far better than the team and reject three times as many people on experience. That is one conversation about screening too hard, not four about effort.
Illustrative
  • It needs rejection reasons

    Which is why they are captured as structured data rather than free text.

  • Against team medians

    A number without a comparison is a number nobody acts on.

  • The same idea runs on sales calls

    Call intelligence coaches reps on talk ratio and objection handling against their own team's medians.

The fair objections

Because performance data is easy to misuse.

Is this a surveillance tool?

It is built as a coaching surface — every figure sits beside the team median, because a number without a comparison is just a number somebody will argue with. If you use it as a league table it will work as one; we would rather you did not.

What does it need from us to be useful?

Rejection reasons, recorded properly. Almost every interesting reading here — screening too hard, losing people on money, the same bar failing every month — comes out of the reason a recruiter picked when they moved somebody.

Does it account for role difficulty?

Not automatically. A recruiter working three executive searches will look slow against someone running volume, and you should read it knowing that. Cut it by role family before drawing conclusions.

Can a recruiter see their own numbers?

Yes — including their own commission book, earned, pending and paid, with a drill-down into what produced each figure. A conversation where both sides read the same screen is a shorter one.

Does the same idea apply outside recruitment?

The sales side has its own version over recorded calls — talk ratio, objection handling, against the team's own medians. Same principle, different corpus.

Book a demo

Look at one recruiter's declines.

The pattern is usually visible in about ninety seconds, and usually fixable.