# Cost per hire, time to hire and quality of hire: three numbers worth tracking > How a small business can measure what a hire costs, how long it takes and whether it worked, with standard definitions, a worked example, a free calculator spreadsheet, and why the popular "a bad hire costs 30% of salary" figure can't be traced. David at SortTheStack, October 13, 2026. Original: https://sortthestack.com/blog/hiring-metrics Three numbers tell you most of what you need to know about how you hire: **what each hire cost, including your own time; how long it took; and whether the person turned out to be the right choice.** You don't need software or benchmarks to track them, just a few minutes per hire and the same definitions each time. The first two are easy. The third is the one that improves your hiring, and the easiest to skip. **Download the calculator:** [hiring metrics (Excel)](https://sortthestack.com/downloads/hiring-metrics.xlsx). Free, no sign-up. ## Cost per hire The standard definition divides everything you spent on recruiting in a period by the number of people you hired. A U.S. standard for the calculation was published by SHRM and the American National Standards Institute in 2012 (ANSI/SHRM 06001.2012), and SHRM's benchmarking lists the kinds of cost it includes: agency fees, advertising and job board fees, job fairs, referral bonuses, travel for applicants and staff, relocation, recruiters' pay and benefits, and recruiting software. Most of those are **external costs**, money paid to someone else. For a large company, the **internal cost** is mostly a recruiting team's salaries. For a small business there is usually no recruiting team: the internal cost is the owner's or manager's time. Leave it out and every hire looks nearly free, which is how small businesses end up underestimating what a slow or repeated search costs them. So count the hours. A simple version: > **Cost per hire** = (external costs + hours spent × hourly value of that > time) ÷ number of hires "Hourly value" is your choice: a salary converted to an hourly rate, or what you'd bill a client for the same hour. Pick one and use it every time. ### A worked example These figures are invented to show the calculation, for the fictional Larkspur Partners hire used in our guides: | Item | Amount | | --------------------------------------------------------- | ------- | | Job board posting | $450 | | Screening software, one job | $149 | | Background check | $60 | | Time: writing the posting and requirements | 3 hours | | Time: screening applications | 4 hours | | Time: phone screens (8 × 20 minutes) | 3 hours | | Time: interviews and work samples (3 finalists, 2 people) | 9 hours | | Time: decision, references, offer | 3 hours | | Value of an hour of the partners' time | $120 | External costs: $659. Time: 22 hours × $120 = $2,640. **Cost per hire: $3,299**, of which 80% is time. That last figure is the useful one. It tells you where to look for savings: in the hours, not the job board. ## Time to hire and time to fill Two measures get confused: - **Time to fill** runs from when you decided to hire (opened the job) to when someone accepted the offer. SHRM's benchmarking defines it this way, in calendar days. It measures how long the seat was empty, which matters to the business. - **Time to hire** usually means the time from when the successful candidate applied to when they accepted. It measures how long a good applicant waited for you, which matters because good applicants are often talking to other employers. Track both, in calendar days. If time to hire is long, look at where people waited: usually between application and first contact, or between the final interview and the offer. The [applicant tracker](https://sortthestack.com/blog/applicant-tracking-spreadsheet) template records the dates you need. ## Quality of hire This is the number that tells you whether your process works, and there is no standard formula for it. Here is a simple version we'd recommend for a small team, built from things you already have: 1. **At 90 days and again at six months, is the person still in the job?** Yes or no. 2. **How are they doing against the requirements you hired them for?** Take the rows of your [candidate scorecard](https://sortthestack.com/blog/candidate-scorecard-template) or [comparison matrix](https://sortthestack.com/blog/how-to-compare-candidates) and rate them again, 1 to 5, now that you've seen the work. Average them. 3. **Knowing what you know now, would you hire them again?** Yes, probably, or no. Write these down for every hire. The value is in what you compare them with: - **The interview ratings.** If the requirements someone scored highest on in the interview are the ones they're now weakest at, that interview question isn't telling you what you thought. - **The source.** If referrals keep working out and one job board's hires keep leaving, that's worth knowing. - **The screening outcome.** Did your best hires come from the top of the pile, or from people you nearly passed over? After five or ten hires you'll have something no benchmark can give you: evidence about which parts of your own process predict success in your own jobs. Be careful with what you conclude from small numbers. Three hires that didn't work out could be chance. Look for patterns over several hires, and treat any single result as a question, not an answer. ## About that "30% of salary" figure You will often read that a bad hire costs "at least 30% of the employee's first-year earnings", usually attributed to the U.S. Department of Labor. We couldn't find the figure in any Department of Labor publication, and the many articles that repeat it don't link to one. Treat it as an unsourced estimate. The cost of a hire that doesn't work out is real, but it varies enormously with the job, how long the problem lasted and how it ended. You can estimate your own: the cost of the repeated search (from the calculator), plus the work that was lost or had to be redone, plus the time you spent managing the problem. That's a figure you can actually use. ## Don't chase benchmarks Published benchmarks for cost per hire and time to fill come from surveys of other organisations, whose jobs and hiring volumes may look nothing like yours, and averages are easily pulled up by a few expensive hires; SHRM itself recommends reading the median when the two differ. A small business hiring a few people a year won't resemble them. The most useful comparison is with your own previous hires. ## Using the calculator The [calculator](https://sortthestack.com/downloads/hiring-metrics.xlsx) is an Excel workbook: - **Cost of one search** adds up external costs and hours, item by item, as in the worked example above. - **Hires** has a row per person hired: costs, hours and four dates (job opened, application received, offer accepted, start date). It calculates cost per hire, the share that was your time, time to fill and time to hire, and the dates of the 90-day and six-month checks, which turn yellow when they're due. - **Summary** gives medians and averages across your hires, and breaks the quality checks down by source and by screening outcome. You set the value of an hour once, on the first sheet. The examples are the Larkspur figures above; delete them when you start. ## Where SortTheStack fits The biggest block of time in most small-business hires is reading applications. SortTheStack reads every application against your requirements and shows the evidence, so you can spend that time on the candidates worth it. How much time that saves depends on your pile, so measure it with the calculator: hours before, hours after. [See how the screening works](https://sortthestack.com/product). ## Sources - SHRM, _Benchmarking: Talent Access_ (definitions of cost-per-hire and time-to-fill), accessed 12 October 2026. - SHRM (2016), _Workforce analytics report_, citing ANSI/SHRM 06001.2012, _Cost-per-hire American national standard_ (2012).