How to measure quality of hire

SIOP’s practitioner review describes quality of hire as a measure without a universally accepted definition. It recommends considering multiple outcomes, context and fairness. That supports documenting your method rather than adopting an unexplained benchmark.^[1]

Write a short measurement brief: role family, outcomes, observation period, data sources, owners and intended use. A permanent employee and a fixed-term project hire may need different outcome definitions. Keep those differences explicit.

Start with a small set of indicators your organisation can collect consistently. The following are options to evaluate, not a prescribed formula or an industry standard. Avoid treating a manager’s overall impression as the complete measure.

Time to fill, cost per hire and interview scheduling time describe the recruitment process. They can help explain workload or delays, but they do not directly show the employee’s contribution after joining.

Keep these measures alongside the outcome review when useful. A shorter time to fill with unchanged employee outcomes is a different result from an improvement in role performance. Neither should be hidden inside a single number.

If you need a combined score, define comparable scales and weights first. One simple illustrative approach is a weighted average of selected outcome ratings. This example is a planning demonstration, not a validated employee-rating instrument.

Suppose a fictional cohort has a role-performance rating of 80/100 and a calibrated manager-feedback rating of 60/100 at the same checkpoint. With hypothetical weights of 75% and 25%, the score is (80 × 0.75) + (60 × 0.25) = 75/100.

The weights sum to 100%. Changing them changes the result: equal weights would produce 70/100. Publish the components and weights with the score. Do not use an arbitrary threshold to make employment decisions.

If an input is missing, report the score as incomplete under this example’s method. Do not silently substitute zero or drop the input, because that changes the meaning of the comparison. Report retention separately, with its own population and checkpoint.

Consider a fictional fintech infrastructure company reviewing backend engineering hires. The manager defines an initial milestone: explain a relevant service flow and complete an approved change with the review and support required for that role. The company’s engineering owners decide what evidence meets the standard.

At the agreed checkpoint, the reviewer records the work observed and the support actually available. Someone who lacked required access has not had the same opportunity as someone who received it on time. Ticket counts or story points alone would not establish that their contribution was better or worse.

For this fintech-infrastructure employee journey, the hiring review includes role criteria and post-interview feedback, while the employee review includes approved work evidence and support gaps. This is an illustrative approach, not a customer result, security clearance or competence certification.

Agree how the authorised team will connect recruitment records to employee outcomes. Use a reliable identifier and a documented mapping rather than assuming that names match. Retain only information needed for the approved purpose, with access and retention rules set by the responsible owners.

Preserve the criteria and post-interview feedback recorded during hiring. Collect post-hire observations at the agreed checkpoints. Do not reconstruct interview evidence from a successful employee’s current reputation.

Give each input an owner and a due date. If records are incomplete, show the coverage and limitation in the report. Connecting records makes analysis possible; it does not automatically make the measures valid.

Compare people in sufficiently similar roles, using the same scales and checkpoints. State the cohort dates, number of eligible hires, number measured and missing inputs. Keep changes in manager, workload, support or role scope visible.

For example, if six of eight eligible hires have a completed review, report six measured and two missing. Do not describe an average based on six as the result for all eight without explaining the missing records. Small cohorts can fluctuate substantially.

An improvement after changing interviews is a signal to investigate. It does not prove that the interview change caused the outcome: onboarding support, candidate mix or business conditions may also have changed. For stronger conclusions, involve someone qualified to design and interpret the analysis.

Review the evidence with recruitment, managers and the responsible People team. Ask whether the role criteria were useful, interview feedback was relevant and the promised support was delivered. The manager’s onboarding plan can help identify support gaps that the recruitment record cannot explain.

Check whether ratings consistently favour particular reviewers, groups or work arrangements, and investigate the measurement process before treating those differences as employee characteristics. Use job-relevant evidence and appropriate expert review.

Agree one process change and the next checkpoint for reviewing it. Keep earlier definitions and results so you can tell whether an apparent improvement came from changed outcomes or a changed measurement method.