Cost per hire and time to hire: measuring both without a system

7 minute read · Updated 2026-08-27

Cost per hire sounds like something that needs software and a HR department. It does not. Two columns in a spreadsheet, filled in as you go, will tell you more about your recruitment than any report, and they will settle the argument about advertising versus agencies with your own figures rather than anyone else's. This guide shows exactly what to record, how to work out cost per hire and time to hire, and what to do with the answers.

What cost per hire actually means

Cost per hire is everything you spent to fill a role, divided by the number of roles filled. The only discipline it needs is deciding once what counts and then being consistent. Include:

• Advertising spend, including any adverts that produced nothing.
• Agency fees, plus VAT if you cannot recover it.
• Job specific costs: a DBS check, a medical, a licence check, a driving assessment.
• Internal time, valued simply. Take the manager's salary, divide by about 1,950 working hours in a year, and multiply by the hours spent. A manager on £45,000 costs roughly £23 an hour, so nine hours of sifting, phone screens and interviews is about £207.
• Any referral bonus paid.

A worked example, using invented round numbers rather than our prices: say you spent £1,200 on advertising in a quarter and filled three roles, one of which also cost a £4,500 agency fee, and the three processes took twenty five manager hours in total at £23. Your total is £1,200 plus £4,500 plus £575, which is £6,275 across three hires, or £2,092 per hire. Strip the agency role out and the other two cost far less. That contrast, in your own numbers, is the whole point of the exercise.

Leave out the salary itself, and leave out training and onboarding beyond the first day. They are real costs, but they belong to employment rather than recruitment, and mixing them in makes the figure unusable for comparison.

Time to hire and time to fill are not the same

Two clocks are worth running, and they answer different questions.

• Time to hire: from the day a candidate applies to the day they accept your offer. This measures your process, and it is the one you control.
• Time to fill: from the day the vacancy was approved to the day the person actually starts. This measures the business impact, and it includes the notice period, commonly one month and up to three for senior or professional roles.

The gap between them is instructive. If time to hire is nine days but time to fill is eleven weeks, your process is fine and your problem is the notice period, so the fix is planning earlier and staying in touch between offer and start date. If time to hire is six weeks, the delay is inside your building: a manager who sifts fortnightly, an interview diary with no slots, an approval step nobody chases.

The spreadsheet: five columns is enough

One row per vacancy, filled in as it happens rather than reconstructed later:

1. Role, site and the date the vacancy was approved.
2. Source and cost of each advert or fee, with the date.
3. Applications received, and how many were genuinely relevant.
4. Key dates: first application, shortlist agreed, interviews held, offer made, offer accepted, start date.
5. Outcome at six months: still employed, left, or a note on why.

The sixth column, the one most people skip, is worth more than the rest combined. A route that produces cheap hires who leave in eight weeks is not cheap. If applications land in your inbox and on a dashboard with the CV attached, as they do here, the dates in column four are mostly already recorded for you and the job is copying them across.

Reading the numbers without fooling yourself

Small employers make three hires a year, not three hundred, so treat every figure as a rough direction rather than a benchmark. Do not compare your cost per hire to a published industry average, because those averages mix graduate schemes, executive search and volume warehouse recruitment. Compare it to your own last three hires and to the cost of the vacancy staying open, which is the arithmetic in our guide to the cost of hiring an employee.

Watch relevance rather than raw volume. Forty applications with two worth interviewing is a targeting problem in the title or the requirements list, not a success. Six applications with four worth interviewing is a good advert.

Then look for the single biggest delay and remove it. In most small businesses it is the same one every time: nobody opened the applications for eleven days. Booking twenty minutes every morning to triage beats every other process improvement available to you.

Using what you learn

Once you have three or four rows you can make decisions instead of guesses. If advertising fills your hourly and administrative roles at a fixed cost, stop paying percentages for them. If one specialist role took two adverts and still needed a search, budget for that properly next year rather than acting surprised. If your best hires come from staff referrals, formalise it. If your time to hire is drifting past four weeks, that is a diary problem you can fix on Monday.

Record it for a year and you will know your own cost per hire by role type, which is the number to take into any conversation about recruitment spend.

Next step: open a spreadsheet, add the five columns and backfill your last two vacancies from your inbox. Then post your current role with the clock running from day one, and see the employer questions if you want to know what we record and what we do not.