Reporting & Insights

01The pipeline funnel

Stage-by-stage candidate counts across every job in an organisation are computed live from the actual pipeline data — no overnight batch job, no stale export. A funnel that's fat at "screening" and thin at "interview" usually means either the screening criteria are too loose (letting through candidates who then fail an interview-stage bar) or too strict relative to what the interview stage actually tests for. Neither diagnosis is available from a single snapshot; it requires watching the funnel shape across several roles.

02CMS tier distribution as a job-description diagnostic

Because CMS scores are comparable across roles — the same formula, the same 0–100 scale — a role whose applicant pool skews heavily toward the Low and Partial tiers is worth investigating on its own terms: is the job description written with unusually strict, narrow requirements that few real candidates meet? Is the sourcing channel reaching the wrong audience for this specific role? A role whose pool clusters in Exceptional and Strong, by contrast, usually indicates either a well-targeted sourcing channel or a job description whose stated requirements match what the actual applicant pool for that kind of role tends to have.

03Time-to-hire, computed honestly

Time-to-hire is calculated from the actual timestamped stage history on hired candidates — the real gap between an application's creation date and the timestamp it moved to "hired" — not from a self-reported estimate. This produces a number that's occasionally less flattering than a rounded estimate would be, but it's the number that's actually defensible when someone asks how it was calculated.

04Using this to make a case internally

A hiring team asked to justify budget or headcount for recruiting tooling is better served by real numbers computed transparently than a polished narrative — a hypothetical claim like "time-to-hire dropped from 34 to 21 days across the last two quarters, computed from timestamped stage history" survives scrutiny in a way a vaguer one doesn't, precisely because anyone can trace it back to the stage history. The reporting exists to be checked, not just presented.

05What the numbers can't tell you

Reporting describes what happened inside a pipeline; it can't tell you whether the underlying job description was well written, whether an interviewer's feedback was actually thorough, or whether a hire who looked strong on every metric available will actually perform well in the role — that last question is precisely the one no hiring analytics system, including this one, has a strong enough track record to answer confidently, and any dashboard implying otherwise is overstating what the data supports.

A funnel and a score distribution are diagnostic starting points for a conversation a human still needs to have, not a conclusion that replaces one. Reading "this role's applicant pool skews Low-tier" as "therefore rewrite the job description" without actually looking at why — wrong sourcing channel? Genuinely scarce skill set? Unrealistic requirements? — turns a useful signal into a wrong-headed shortcut.