Sourcing KPI Dashboard: Metrics That Help Senior Sourcers Improve Searches
SourcingOS Editorial · Updated August 19, 2026
A practical sourcing metrics guide for lane yield, evidence-fit lead yield, rejection patterns, response, search coverage, Unique Contribution Rate, and recruiter time.
Measure whether the search is improving, not whether the sourcer is busy.
Profiles viewed, names added, and messages sent are activity counts. They can be useful operationally, but they do not tell you whether a sourcing strategy is expanding coverage, finding different people, or learning from rejection.
A stronger sourcing metric set
Evidence-fit lead yield
How many leads survive the same job-relevant review standard per lane or source?
Unique Contribution Rate
What share of a source’s evidence-fit leads did none of the other tested sources surface in the same requisition-level test?
Duplicate pressure
How much of recent sourcing work is resurfacing people the team already knows?
Rejection pattern
Which evidence gaps or hidden standards repeatedly cause hiring-manager rejection?
Response by outreach angle
Which evidence-based outreach framing produces replies without collapsing into high-volume messaging?
Recruiter time
How much time does each lane require after review, correction, deduplication, and handoff are included?
Keep UCR in context
Unique Contribution Rate measures additive discovery. It does not replace response, downstream outcomes, cost, recruiter time, or role-family context. Keep the raw counts next to the rate.
Use metrics to choose the next search action.
If duplicate pressure is rising but an untested donor-company lane remains, open the lane. If adjacent-title yield is still producing evidence-fit leads, calibrate before declaring the market exhausted. If a source has low unique contribution and high workflow cost across repeated comparable tests, that becomes a stack decision worth reviewing.
Explore Candidate Search — Carry the role requirements into candidate discovery and inspect evidence before deciding.