Recruitment Tracker: Build an Evidence-Ready Workflow

A recruitment tracker should answer a practical question in seconds: what needs attention next, and what evidence supports the current state? Too many trackers answer only where a candidate sits. A cell says “screen,” “interview,” or “review,” but does not show what has been verified, what remains unclear, or who owns the follow-up.

That gap matters when recruiters manage several roles, hiring managers review candidates at different times, or evidence moves between resume review, phone screens, and interviews. The tracker becomes useful when it carries the decision context forward without pretending to make the decision itself.

This guide shows how to build a recruitment tracker around stages, evidence, owners, and exceptions. It is an implementation layer for an existing recruitment pipeline, not a second process competing with it.

Define what the tracker is responsible for

A recruitment tracker is a shared operational record for active roles and candidates. It should show the current workflow state, the evidence available for the next review, unresolved questions, and the person responsible for the next action.

Keep three responsibilities separate:

1. Pipeline design: the stages, entry rules, exit rules, and allowed transitions.
2. Candidate tracking: the candidate's current state, relevant dates, action owner, and next step.
3. Candidate evaluation: the evidence and human judgment recorded against approved role criteria.

A stage field cannot replace an evaluation. “Interview” does not explain why the candidate advanced. “Hold” does not identify the missing evidence. “Strong” does not name the criterion or source. The tracker should link these layers while preserving their boundaries.

Current recruiting-system documentation illustrates the operational distinction. Greenhouse's Visual Candidate Pipeline documentation says its view summarizes active candidates in each stage of your job's pipeline, shows actions that need attention, and exposes days in stage. Those are tracker signals: location, age, and pending work. They help a recruiter direct attention, but they are not candidate evidence on their own.

Start with the minimum useful field set

Begin with fields that change a review or handoff. A compact tracker can support a disciplined workflow if every field has a clear purpose.

Role fields

• Role ID and title: stable identifiers that prevent records from being mixed across similar searches.
• Hiring manager and recruiter owner: the people accountable for role decisions and workflow movement.
• Criteria version: the approved role definition used when the candidate was reviewed.
• Pipeline version: the stage structure and transition rules currently in force.

Candidate workflow fields

• Candidate ID and source record: a stable identity plus a link to the resume, profile, or submitted record under review.
• Current stage: one value from the approved pipeline.
• Stage-entry date: when the candidate entered the current stage.
• Next action and owner: the specific task and the person responsible for it.
• Action due state: ready, waiting on internal review, waiting on candidate input, or blocked by an identified exception.
• Last meaningful update: the most recent evidence, decision note, or workflow change rather than a generic record edit.

Evidence fields

• Evidence status: ready for review, incomplete, conflicting, or source check required.
• Scorecard status: not started, in progress, submitted, or needs clarification.
• Open question: one concise question that could change the next decision.
• Decision note: the reason for the latest human-owned advance, pause, or decline, tied to job criteria.

Keep the full source material outside the summary row. Link to the verified resume record, interview notes, and scorecard so a reviewer can reconstruct the reasoning. The guide to resume parsing quality control explains how to keep extracted fields connected to the original document before they enter screening.

Use controlled states instead of improvised labels

Tracker language should be boring and precise. Define the allowed values for stage, evidence status, next-action type, and exception reason. If one recruiter writes “pending,” another writes “waiting,” and a third writes “HM review,” the team cannot reliably filter the work or understand why it stopped.

Write a short definition for every value. “Evidence incomplete” might mean at least one required criterion lacks enough source support for a screening conclusion. “Source check required” might mean the structured record and original resume disagree. “Scorecard needs clarification” might mean a rating was submitted without a supporting example.

Do not use labels that collapse uncertainty into a verdict. Replace “weak candidate” with the criterion and state: “Required implementation ownership not established in the resume; verify scope if the recruiter advances to a conversation.” That wording tells the next reviewer what is missing without claiming the candidate lacks the capability.

Make every stage change carry evidence

Require four items when a candidate changes stage: the new stage, the transition reason, the evidence source, and the human owner. The reason can be brief, but it should name the job-related criterion or workflow event that made the transition appropriate.

For example: “Advance to recruiter screen. Resume shows ownership of two multi-location system rollouts; confirm project scale and the candidate's direct responsibilities.” The note preserves both supporting evidence and the unresolved question. A later reviewer can see why the candidate moved and what the next stage must examine.

Use the same structure for pauses. “Awaiting hiring-manager review of required domain depth” is more actionable than “on hold.” If the process cannot continue because a file is unreadable, a scorecard is incomplete, or two records appear to conflict, route the candidate to an exception state with a named owner instead of changing the candidate's evaluation.

Connect the tracker to scorecards without duplicating them

The tracker should show whether evidence is ready, not reproduce every rating. Greenhouse's scorecard overview describes scorecards as a way for a hiring team to review candidates using predetermined criteria and submit interview feedback against those criteria. That makes the scorecard the evaluation record; the tracker should point to it and show its completion state.

Use one candidate scorecard definition for everyone reviewed for the role. In the tracker, record which evaluation is expected at the current stage, whether it is complete, and whether any material criterion still lacks evidence. Avoid copying a total rating into the tracker without the criteria and source context that produced it.

Create views for action, not decoration

One source record can support several focused views. Build each view around a question a recruiter or hiring manager needs to answer:

• Recruiter action queue: candidates with an internal next action, grouped by owner and stage age.
• Evidence exception queue: candidates whose source record, scorecard, or criterion evidence needs human review.
• Hiring-manager review queue: candidates ready for a defined decision, with the evidence link and open question visible.
• Stale-state view: records whose stage has not changed and whose next action lacks a current owner or explanation.
• Role calibration view: repeated open questions or disagreement patterns that may signal an unclear role criterion.

Do not treat age alone as failure. A candidate can remain in a stage for a known reason. Pair time in stage with action state and owner so the team can distinguish purposeful waiting from abandoned work.

Set update rules and review ownership

Decide which event updates each field. A completed resume review may change evidence status and next action. A submitted interview scorecard may change scorecard status but should not automatically change stage. A recruiter decision may move the candidate only after the required evidence is available.

Name who can make material changes and who checks exceptions. If automation proposes an update, record the source event and keep the change provisional until the appropriate recruiter reviews it. The system may summarize notes, suggest an action owner, or flag an empty field. It should not infer a rejection, invent evidence, or advance a candidate because a deadline elapsed.

Audit a small sample of tracker rows against source records and scorecards. Look for unsupported labels, stale owners, missing decision notes, stage changes without reasons, and open questions that disappeared without resolution. Use the patterns to improve field definitions or handoff rules.

Measure whether the tracker supports better review

Use the tracker to diagnose workflow, not to manufacture a performance story. Useful checks include the share of active candidates with a named next owner, the share of stage changes with a decision note, unresolved evidence exceptions, overdue scorecards, and time between a material question and its resolution.

Pair counts with examples. If one role has repeated “criteria unclear” exceptions, inspect the role definition. If interview feedback often arrives without evidence, review the scorecard instructions. If stage age rises because the next owner is blank, fix the handoff. The recruiting metrics guide shows how to connect a signal to a workflow correction instead of treating the number as a verdict.

Recruitment tracker quality checklist

• Does every active candidate have one current stage?
• Does the stage use a controlled value from the approved pipeline?
• Is the stage-entry date visible?
• Does every pending action have a named owner?
• Can the reviewer open the relevant source evidence?
• Are missing evidence and confirmed misses kept distinct?
• Does every material stage change include a reason?
• Does the tracker point to the scorecard instead of flattening it?
• Are exceptions visible without changing the candidate's evaluation?
• Are AI-proposed updates checked by a human?
• Can a skeptical reviewer reconstruct the latest decision?

Build the tracker around the next responsible action

A recruitment tracker earns trust when it does more than display candidate locations. Give every record a current stage, an evidence state, a specific next action, and a human owner. Link summary fields to the source and scorecard, preserve open questions, and require reasons for material changes. The result is a tracker that helps recruiters move work forward while keeping hiring judgment evidence-based and human-owned.