Quality of hire metrics: how to measure, score, and act on it

Why quality of hire is the metric everyone wants and few get right

Ask a room of talent leaders which recruiting metric matters most and you will hear "quality of hire" almost in unison. Ask the same room what their current quality of hire score is, and the answers go quiet. Harver's industry survey lands on the gap squarely: 39 % of talent leaders rank quality of hire as their top performance priority, only 33 % claim a strong measurement approach, and a tiny 5 % call themselves best-in-class. SHRM, less subtly, calls it the holy grail of recruiting.

The reason is unflattering. Quality of hire metrics ask the awkward question: did the people you hired turn out to be the people you needed. Cost-per-hire and time-to-fill are easier; the data falls out of an applicant tracking system without anyone having to form a view. Quality of hire requires you to look back and grade your own work. Most teams measure what is convenient, not what is useful.

There is a familiar rationalisation here - "we know a good hire when we see one" - and it is exactly the signal the metric is meant to replace. If your gut were that calibrated, you would not need a recruiting function. The point of a quality of hire metric is to stop the firm from re-running the same interview-then-vibe loop and calling it a process.

This piece walks through what the metric actually means, the formula HR teams use, the scorecard that produces honest data, and the post-hire signals worth tracking. It also addresses the strongest critiques - because they are real, and ducking them produces a quality of hire KPI nobody trusts.

What quality of hire actually means

The cleanest definition is the dullest one, and it comes from the standards body. ISO/TS 30411:2018 defines quality of hire as the performance of an individual after hire compared to pre-hire expectations. Two halves matter: the performance, which you observe later, and the expectations, which you set at decision time. A quality of hire metric without both halves is a vibe with a number on it.

It helps to be clear about what quality of hire is not. Time-to-fill measures how quickly you closed the requisition. Cost-per-hire measures what you spent doing it. Offer-acceptance rate measures whether the candidate said yes. All three are about how you hired. Quality of hire is about who you hired. They are not interchangeable, and treating them as one dashboard is how recruiting teams end up optimising for speed and cost while quietly accumulating hiring debt.

Quality of hire is also, by design, a composite. There is no single observable that captures whether the person you hired was the person you needed - so every company assembles its own index from a handful of inputs. That flexibility is why the formula varies, and why the choice of inputs matters more than the arithmetic.

Asked plainly, the quality of hire KPI is about one thing: did the people you hired turn out to be the people you needed. Everything that follows is the apparatus for answering that question with evidence rather than opinion.

The quality of hire formula and how to calculate it

The most cited quality of hire formula is also the most defensible. It averages three indicators that any HR function can reasonably collect: performance rating, productivity ramp, and one-year retention. Written out:

QoH index = (PR + HP + HR) / 3

PR is the average performance rating, productivity ramp, and one-year retention of new hires expressed as a percentage. HP is the percentage of new hires who reached an acceptable productivity level inside the agreed window - typically 90 days for individual contributors, longer for managers. HR is the percentage of new hires retained at the 12-month mark. Each is a number between 0 and 100; the average is your quality of hire calculation, expressed the same way.

A worked example helps. Suppose you hired 30 people last year. The cohort's average performance rating after their first review cycle is 76 % of the maximum on your scale. 24 of the 30 hit their productivity ramp on time, giving HP = 80 %. 27 are still with you at 12 months, giving HR = 90 %. Quality of hire = (76 + 80 + 90) / 3 = 82 %. That is your headline number, calculated quarterly or yearly and trended over time.

The general form is simply the average of any N indicators you decide matter. Some teams add a hiring-manager satisfaction score; others add new-hire engagement; a few weight the inputs unequally. The trick is choosing inputs you can actually collect every quarter without a special project. A QoH formula that requires bespoke data extraction once a year is a formula you will run once.

One caveat on the retention input. Around 20 % of new hires leave within the first 45 days, so retention windows shorter than three months overstate the metric. A 30-day retention figure tells you something about onboarding shock; the 12-month figure is the one that correlates with whether the hire was right.

The scorecard that produces the data

A quality of hire scorecard is, in plain terms, a one-page list of capabilities with anchored ratings against the capabilities the role demands. It is the artefact that turns an interview into data. Without it, you are recording impressions; with it, you are recording evidence the QoH index can later be checked against.

This is not a clever HR-tech invention. It is half of the framework Campion, Palmer and Campion (1997) laid out in their review of the 15 components of interview structure. They split structure into two sides: the content of the interview - questions derived from job analysis, asked the same way of every candidate, no helpful prompts, candidate questions held until the end - and the evaluation discipline - rating each answer on an anchored scale, multiple raters, consensus scoring, interviewer training, and statistical decision rules. The scorecard sits on the evaluation side. Get it right and the validity uplift the meta-analyses keep finding becomes available to you.

Anchor each capability to behaviours, not opinions. This is McClelland's (1998) Behavioral-Event Interview principle in scorecard form: capabilities are most reliably identified through what people did in past situations, not through how they describe themselves. A scorecard row reading "communication: 1-5" is not anchored. A row reading "explained a technical concept to a non-technical stakeholder; check whether they adapted vocabulary, structured the explanation, and confirmed understanding" is.

A 1 to 5 anchored scale, where each number has a written description of what that level of answer looks like, beats a 1-to-10 vibes rating every time. It also makes inter-rater agreement an actual measurement rather than a polite request. Two interviewers landing on a 4 and a 2 for the same answer is a useful disagreement; two interviewers landing on an 8 and a 6 is statistical noise.

The discipline gap is where most scorecards fail. The scorecard works only if interviewers fill it in independently before they discuss the candidate. Pooled judgement during the room - the panel turning to the senior person and asking what they think - is the failure mode. If the structured interview ends with a debate and one shared score, you have a paper exercise, not a measurement. The data flowing into your QoH calculation needs to come from interviewers who scored alone, then compared.

Post-hire signals: hiring manager surveys, ramp, retention

The most useful post-hire signal is also the cheapest. A short hiring-manager survey at 30, 60, and 90 days after the new hire starts - a 1-to-5 fit rating plus two open-ended questions - produces a leading indicator that arrives in time to act. By month three you know whether the hire is working out. By the time the 12-month retention number lands, the answer is forensic.

Pair the manager survey with a new-hire fit rating, where the new joiner answers the inverse question: does the job match what was described in the interview process. The two views together catch the failure modes a one-sided survey misses. A manager who says "fine" while the new hire says "not what I was sold" is information; either signal alone is not.

Ramp time and 12-month retention complete the standard four-input composite that most teams should start with. Score from the structured interview goes in the front. Performance rating, ramp, retention, and the manager-survey average come out the back. That is enough inputs to triangulate without becoming a research project. ClearCompany's catalogue of the eighteen metrics that feed quality of hire is a useful menu, but pick four to six and stick with them. A QoH index built on twelve inputs is a QoH index nobody believes.

This is also where the question of predictive validity becomes concrete. The reason to collect post-hire data is not just to grade the cohort; it is to check whether your scorecard scores actually correlate with later performance. If interviewers gave a candidate a 4.6 and the manager survey six months later says "below expectations", the scorecard item that produced the 4.6 is not earning its place. Without post-hire data the scorecard is a hopeful artefact; with it, the scorecard becomes self-correcting.

How to install a measurement system that survives contact with reality

Here is the part nobody puts in the formula post. Quality of hire metrics only work when the interview underneath them is structured. McDaniel et al. (1994) reviewed 245 validity coefficients across 86,311 individuals and put structured interviews near the top of any predictive-validity ranking that mattered; Campion et al. (1997) reported corrected validities of .35 to .62 for structured interviews against .14 to .33 for unstructured. Levashina et al. (2013) refreshed the picture twenty years on and reached the same conclusion across twelve subsequent meta-analyses. Without that uplift, the QoH index is just measuring whether your gut feel got lucky, then averaging the result.

This is the gap a self-guided programme is designed to close. HireSchool sells the Structured Hiring Method - a self-guided digital programme delivered as video content plus a learning management system. You buy access, your team works through the curriculum, and the method gets installed inside your business rather than rented from a consultancy. It codifies the components the research says do the work: leadership values translated into capabilities, codified scorecards with anchored ratings, behavioural interviewing training, decision management, and an optional quality-assurance module that ties post-hire ratings back to the scorecards that produced them.

Two pieces of the programme matter most for quality of hire metrics specifically. The first is the codified scorecard. Every interviewer scores the same candidate against the same anchored capabilities, independently, before the panel discusses anything. That is the part of structure that does the validity work, and it is the input your QoH index needs to be measuring against. The second is the QA loop. Post-hire signals - the manager survey, ramp time, retention - feed back into which scorecard items predicted performance and which did not. Items that consistently fail to predict get rewritten or removed; items that predict well stay. The QoH index becomes a learning system rather than a year-end report.

HireSchool is not consultancy, not an applicant tracking system, and not a recruiting agency. There is no embedded HireSchool team running the process inside your business. The programme teaches your hiring managers and interviewers how to run it; you keep the muscle when the work is done. That is deliberate - the people making hiring decisions are the people who need the discipline, and outsourcing the discipline keeps producing the same hiring debt.

If you want to see what an installed quality-of-hire system actually looks like before the post-hire data starts arriving, explore the Structured Hiring Method programme. The scorecards, interview flow, and decision mechanics are the artefacts your QoH index will eventually be evaluating. Putting them in place first is what separates a metric you can defend from a metric that drifts every year because nobody can quite remember why it moved.

Honest objections and where the metric still goes wrong

The strongest critique of quality of hire metrics is that they are lagging indicators. By the time the score is in, the bad hire has already cost you weeks of ramp, a rework, and the morale of the team that absorbed the gap. Matt Charney puts it sharper, calling QoH a vibe dressed up as a KPI. He has a point. The answer is not to abandon the metric; it is to instrument the leading indicators - scorecard scores at decision time, manager 30-day surveys - that arrive in time to act, and treat the 12-month index as the audit, not the alarm.

The subjectivity problem is real too. Performance ratings are themselves imperfect; the QoH score is only as reliable as its inputs. This is why the structured interview underneath matters as much as the formula on top. Dana, Dawes and Peterson (2012) showed that even random unstructured interview information dilutes the predictive value of better evidence; Kausel, Culbertson and Madrid (2016) showed it inflates interviewer confidence. Without structure, your QoH numerator is contaminated, and a contaminated index trends however the noise feels like trending that quarter.

A brief word on the related question of the 5 Cs of hiring - character, competence, culture, chemistry, capacity, depending on whose list you read. They are useful as a cultural frame for what you are looking for. They are not a measurement system. Treat them as adjective territory and translate the ones that matter into anchored scorecard items, or you will end up scoring "chemistry" on a five-point scale and pretending the result means something.

The handoff is clean. A working metric needs a working scorecard. A working scorecard needs a structured interview. The rest is bookkeeping.