The quality of hire formula, worked in full

There is no single accepted quality of hire formula. What exists is a family of approaches that combine a handful of components — usually a performance or manager satisfaction rating, a retention or tenure measure, and sometimes a ramp-time score — into one number, weighted according to what a given organization decides matters most. This page shows three of those approaches with the arithmetic worked all the way through on invented numbers, so you can build your own rather than hunting for a formula that does not exist in a standardized form.

For the broader question of what to measure and when, see how to measure quality of hire, which covers choosing components and running the manager survey this page's arithmetic depends on.

Why there is no single formula

Quality of hire is not a directly observable quantity the way revenue or headcount is; it is a construct built from proxies (manager opinion, a performance rating, whether someone stayed) that each capture part of the picture and none of it perfectly. Different organizations weight those proxies differently because they value different things: a sales organization cares enormously about ramp time and quota attainment, while a research team might weight long-term retention and peer assessment more heavily. That is a reasonable disagreement, not a gap someone forgot to fill with a standard. The goal below is not to find the "real" formula but to build one your organization can defend and keep using consistently.

The building blocks most formulas use

Formula A: a simple weighted average

Rescale every component to a 0-100 scale, assign a weight to each that sums to 100 percent, and take the weighted sum. This is the easiest formula to explain to a hiring manager who has never seen the number before.

Quality of hire = (w1 x manager satisfaction) + (w2 x performance rating)
                + (w3 x retention) + (w4 x ramp time)
where w1 + w2 + w3 + w4 = 1.0

Example weights: manager satisfaction 30%, performance rating 40%,
retention 20%, ramp time 10%

Worked example (invented hire, "Casey R."):
- Manager satisfaction: survey average 4.2 out of 5 -> 4.2 / 5 x 100 = 84
- Performance rating: 3.9 out of 5 on the standard review -> 3.9 / 5 x 100 = 78
- Retention: still employed at the 12-month mark -> 100
- Ramp time: hit full productivity 5 days ahead of the expected 90-day target -> 90

Quality of hire = (0.30 x 84) + (0.40 x 78) + (0.20 x 100) + (0.10 x 90)
                = 25.2 + 31.2 + 20.0 + 9.0
                = 85.4

The number 85.4 means nothing in isolation. It becomes useful only when compared against the same formula applied to other hires, which is what Formula B does.

Formula B: an index against your own cohort average

Instead of reading the weighted score on its own, index it against the average score of a comparable cohort — the same role family, hired in the same period, or the same source. This turns an unanchored number into a comparison, which is closer to how the score actually gets used in practice.

Quality of hire index = (hire's weighted score / cohort average weighted score) x 100

Worked example, continuing Casey R.:
- Casey's weighted score (from Formula A): 85.4
- Average weighted score across the last 10 hires in the same role family: 76.0

Quality of hire index = (85.4 / 76.0) x 100 = 112.4

Reading it: Casey scored about 12% above the trailing average for this role family.

This version answers a more useful question than Formula A alone: not "is 85.4 good," which has no fixed meaning, but "is this hire better or worse than what this role family has produced recently," which a hiring manager can actually act on.

Formula C: a retention-adjusted formula

Formulas A and B treat retention as one component among several, which understates how much a fast departure costs. This version multiplies a blended performance-and-satisfaction score by a retention multiplier, so a strong early performer who leaves within a few months scores low overall, reflecting that the hiring investment was not recovered.

Quality of hire = (blended performance-satisfaction score) x (retention multiplier)

Blended score = average of manager satisfaction and performance rating, both on 0-100
Retention multiplier:
  left within 6 months voluntarily      = 0.0
  left between 6 and 12 months          = 0.5
  retained past 12 months               = 1.0

Worked example (invented hire, "Dana L.", who left at month 9):
- Manager satisfaction: 90 (survey average 4.5/5 x 100)
- Performance rating: 85
- Blended score = (90 + 85) / 2 = 87.5
- Retention multiplier for leaving at month 9 = 0.5

Quality of hire = 87.5 x 0.5 = 43.75

Dana's blended score before the retention adjustment (87.5) looked like a strong hire. The retention-adjusted score (43.75) tells a different story: whatever made Dana leave at month nine is worth investigating regardless of how good the early performance signal was, because the formula is now treating the departure as a real cost, not a footnote.

Choosing weights for your organization

Start from what you actually want the score to predict. If the main risk you are managing is early attrition, weight retention and Formula C's multiplier structure more heavily. If the main risk is a slow-to-ramp hire quietly underperforming for months before anyone notices, weight ramp time and manager satisfaction more heavily and check them earlier than 90 days.

Test your chosen weights against a small set of hires everyone already agrees were clearly strong or clearly weak. If the formula produces a mediocre score for someone everyone remembers as an excellent hire, the weights are wrong, not the hire. Adjust the weights until the formula matches reality for the cases you are confident about, then trust it for the harder-to-judge cases in the middle.

A full worked example, start to finish

All numbers below are invented for a fictional customer support hire, "Morgan T.," to show every step in one place.

What the number is useful for, and what it is not

Use the score to compare hiring sources, teams or time periods in aggregate — for example, whether hires from a particular sourcing channel trend higher or lower than hires from another over a year. That is a reasonable, defensible use because it is comparing groups, where the noise in any one person's score washes out. Do not use a single hire's score as the sole input into a performance or termination decision about that person; the formula is built from proxies with real error in them (a manager's mood on the day they filled out a survey, a performance rating cycle that ran differently that quarter), and treating it as a precise individual measurement gives it more authority than the arithmetic behind it can support.

It is also worth being explicit with hiring managers about what the score is not measuring: it says nothing about whether the interview process itself was good, only about what happened after the hire started. A role with a low average quality-of-hire score might point to a screening problem, a management problem, or a role that was poorly scoped from the start, and the formula alone cannot tell you which.

Common arithmetic mistakes

The most common mistake is mixing scales without rescaling first — averaging a 1-5 survey score directly with a 0-100 performance score produces a meaningless number dominated by whichever component happens to have the larger range. Always rescale every component to the same range before applying weights. A second mistake is letting weights silently sum to something other than 100 percent after a component is added or removed; recheck the sum every time the formula changes. A third is treating a single cohort average as permanent; recompute it each time you score a new group, since a cohort average built from three hires is not a stable baseline the way one built from thirty is.