Process and pipeline · 6 min read

Hiring metrics that matter

A short set of well-defined hiring metrics tells you where your process is slow, where good people drop out and whether your hires work out. This guide explains what to measure and exactly how to calculate it.

Measure to make decisions

Hiring metrics are worth tracking only if they change what you do. Before adding a number to a report, ask what you would do differently if it went up or down. If the answer is nothing, leave it out. A short set of metrics that the team reviews every month is worth more than a dashboard nobody opens.

This guide gives you the definition and the calculation for each metric. It does not give target figures. Roles, markets and processes differ too much for an outside number to mean anything. Your own trend over time is the comparison that matters: are you getting faster, are more of your hires working out, are fewer good people dropping out.

Before you start, agree on dates. Decide what counts as the start of a search, such as the day the role is approved or the day the job ad goes live, and what counts as the end, such as offer accepted or start date. Write the definition down and use it every time.

Speed metrics

  • Time to hire: the number of days from when an applicant enters your pipeline to when they accept the offer. Calculate it for each hire, then take the median across all hires in the period. The median is less distorted by one unusual hire than the average.
  • Time to fill: the number of days from when the role is approved to when an offer is accepted. This includes the time before anyone applies, so it shows delays in approvals and in writing the job ad.
  • Time in stage: the number of days an applicant spends in each pipeline stage. Calculate the median for each stage separately. This is the most useful speed metric, because it tells you exactly where the delay sits.

Time to hire and time to fill are headline numbers. Time in stage is the one you act on. If the median wait between final interview and offer is long, look at approvals. If the wait between application and first screen is long, look at screening capacity. For ways to cut delay once you know where it is, see the guide on reducing time to hire.

Pipeline health metrics

  • Applicants per opening: total applicants divided by the number of openings. On its own it says little. Read it alongside the qualified applicant rate.
  • Qualified applicant rate: applicants who meet the must-have requirements, divided by total applicants, multiplied by 100. A falling rate suggests the job ad is attracting the wrong people or the requirements are unclear.
  • Stage conversion rate: applicants who moved from one stage to the next, divided by applicants who entered that stage, multiplied by 100. Track it for every stage.
  • Withdrawal rate by stage: applicants who withdrew at a stage, divided by applicants who entered that stage, multiplied by 100. Withdrawals clustered at one stage point to a problem there, such as a long wait or a heavy task.
  • Offer acceptance rate: offers accepted divided by offers made, multiplied by 100. Record the reason for every declined offer, because the reasons matter more than the rate.

Read these together. A stage where conversion is close to total may not be filtering anything and could be dropped. A stage where conversion is very low may be catching problems that should have been caught earlier, at the CV screen.

Source metrics

  • Applicants by source: a count of applicants from each channel, such as your careers page, job boards, referrals, agencies or your talent pool.
  • Qualified applicants by source: the same count, only for applicants who passed screening.
  • Hires by source: the same again, only for people hired.
  • Cost per hire by source: what you spent on a channel in the period divided by hires from that channel. Include advertising, agency fees and referral bonuses. Leave out the salaries of the hiring team unless you include them for every source.

Volume and quality often point in different directions. A channel that sends many applicants and few hires costs screening time. A channel that sends few applicants and many hires deserves more attention. To count sources reliably, record a source for every applicant at the point they apply. Tagging the link they used is more reliable than asking them.

Quality metrics

Speed and volume mean little if the hires do not work out. Quality is harder to measure, but you can get a usable signal.

  • Hiring manager rating: at a fixed point after the start date, such as the end of probation, ask the hiring manager to rate the hire against the scorecard used at interview. Use the same scale every time.
  • Early leaver rate: hires who leave within a set period after starting, divided by total hires in the same group, multiplied by 100. Choose the period once and keep it.
  • Probation pass rate: hires who complete probation divided by hires who reached the end of it, multiplied by 100.
  • Candidate experience score: a short survey sent to applicants after they leave the process, hired or not. Ask one rating question and one open question. The guide on candidate feedback covers how to run it.
  • Hiring manager satisfaction: a short rating from the hiring manager after each search closes, on the quality of the shortlist and the speed of the process.

Quality metrics take months to arrive. That is fine. Review them every quarter rather than every week, and compare groups of hires rather than single people.

Set up and review your metrics

  1. Pick four to six metrics. A sensible start is time in stage, stage conversion, offer acceptance, hires by source and hiring manager rating.
  2. Write a definition for each one, including the start and end dates and what to include or leave out.
  3. Make sure the data is captured as part of normal work: stages moved in one system, a source on every applicant, a reason on every rejection and withdrawal.
  4. Review the numbers monthly with the hiring managers who rely on them.
  5. For each metric that moves the wrong way, agree one change and check its effect the following month.

If your hiring data lives in spreadsheets and email threads, most of these numbers will be hard to collect. The guide on moving hiring off spreadsheets covers how to fix that first. Spark Careers Enterprise includes reports and CSV export of applicant and stage data, so you can calculate these metrics from records the team already keeps instead of rebuilding them by hand.

Common mistakes to avoid

  • Comparing your numbers with figures from elsewhere instead of with your own past.
  • Changing a definition partway through the year.
  • Using averages when one slow hire distorts the picture. Use the median.
  • Measuring speed without measuring quality.
  • Tracking a metric nobody acts on.
  • Using metrics to blame individuals rather than to fix the process.
  • Collecting personal data for reporting without checking the employment and data protection rules where you hire.

Your next hire, start to finish, in one place.

Free workspace. No card. Full scoring free for 7 days.