Three numbers get asked for in every hiring review. Two are straightforward if you define them carefully. The third cannot really be measured, and pretending otherwise is the most common mistake in hiring reporting.
Time to hire and time to fill are different
They answer different questions and get used interchangeably, which is why nobody trusts either.
| Time to hire | Time to fill | |
|---|---|---|
| The clock starts | When a candidate enters the process | When the job is approved |
| The clock stops | When they accept | When they accept |
| It measures | How fast your process moves a person | How long the business waits |
| Who owns it | Recruiting | Recruiting and the business together |
Report both, clearly labelled. Time to hire is what you can improve directly. Time to fill is what your colleagues actually feel, and it includes the two weeks before anyone could start searching because nobody had approved the job description.
Use the middle value, not the average. One nine-month search wrecks an average and tells you nothing about the normal experience. Report the middle value with the range beside it.
Cost per hire
Include what you actually spend finding people:
- Job advertising
- Agency fees
- Software, spread across the hires it helped with
- Referral bonuses
- Recruiter time, if you can work it out honestly
Leave out — or report separately — salary, relocation, signing bonuses and equipment. Those are costs of employing someone, not of hiring them. Mixing them in makes senior jobs look like recruiting failures.
One important warning: cost per hire only makes sense between similar jobs. An agency-filled executive search and a high-volume support role are different businesses. A single company-wide figure moves whenever your mix of jobs moves, and people will read that as a change in performance.
Quality of hire
There is no clean way to measure this. Every option is a stand-in with a known flaw. The honest approach is to use two or three and say what is wrong with each, rather than build one number that hides the flaws.
| Stand-in | Its flaw |
|---|---|
| Performance rating at 6 or 12 months | Ratings vary by manager and carry their own biases |
| Still there after 12 months | Slow, and affected by the manager and team more than the hire |
| Time to being fully productive | Rarely measured consistently; depends on onboarding |
| Hiring manager satisfaction | Measures the relationship as much as the hire |
| Passed probation | A low bar. Nearly everyone does |
The trap to avoid: combining these into a single "quality score" with weights nobody questions. It creates the appearance of measurement while making the flaws invisible — and it will be used to judge people, both recruiters and hires, on a number that cannot carry that weight.
Presenting all three honestly
- Show the count behind every percentage.
- Use middle values and ranges, not averages.
- Split by type of job. Refuse to produce one company-wide number when the mix varies.
- Name the stand-in. "Still employed at 12 months, as a rough measure of quality" is a sentence that prevents a lot of later misuse.
- Say what a number cannot tell you. It builds more credibility than any dashboard.
The pressure to watch for
Quality-of-hire numbers get requested when someone wants to judge the recruiting team. Agreeing to a number you know is weak, because it is what was asked for, is how a team ends up being managed against noise.