Cost per hire, and the screening hours inside it

For one 3,000-applicant drive in our pilot, HR’s manual profile review would have taken about 18 hours. With AI screening it took 1-2 hours of dashboard review. Those 18 hours were real money, yet in a typical cost per hire report they never appear as a line of their own. The job board invoice did. The agency fee did. The recruiter reading resumes until 9 PM did not.

That is the problem with the metric as most teams run it. The external costs are easy to find because someone sends a bill. The internal ones, mostly people’s time, get estimated, averaged or left out. In high-volume hiring, the time is the biggest cost there is.

What is cost per hire?

Cost per hire is the total amount spent on recruiting in a period, internal and external, divided by the number of hires made in that period. Internal costs are your own team’s time and tools; external costs are what you pay outside parties, such as job boards, agencies and background checks.

The formula is standard. In 2012 the American National Standards Institute accepted a cost-per-hire standard developed by SHRM, and the coverage at the time noted that recruiter salaries were already part of typical calculations. Written out:

Cost per hire = (total internal recruiting costs + total external recruiting costs) ÷ total number of hires

The arithmetic is easy. Deciding what goes into each bucket is where most of the error comes from.

What counts as an internal or external cost

AIHR’s cost per hire guide lists the usual items, and it is a useful checklist.

Two rules make the number honest. Count the same items every period, or your trend line measures your accounting, not your hiring. And count time as a cost even though nobody invoices it.

A worked example in rupees

Here is a hypothetical role to show the arithmetic. The numbers are illustrative, not a benchmark: replace each one with your own.

A company is hiring 20 customer support executives in one month. It gets 1,500 applications.

External costs

• Job portal listings and boosts: ₹75,000
• Background checks, 20 × ₹1,000: ₹20,000
• External total: ₹95,000

Internal costs (time valued at an assumed ₹500 an hour, salary plus overheads)

• Reading 1,500 resumes at about 1 minute each: 25 hours
• Phone screens for 400 shortlisted candidates at 12 minutes each, including dialling and notes: 80 hours
• Hiring manager interviews for 60 candidates at 30 minutes each: 30 hours
• Total: 135 hours × ₹500 = ₹67,500
• Internal total: ₹67,500

Cost per hire = (₹67,500 + ₹95,000) ÷ 20 = ₹8,125

Now look at where the money went. The resume reading and phone screens alone are 105 hours, or ₹52,500. That is about a third of the entire cost of hiring these 20 people, and it is the part least likely to appear in a cost per hire report. The market cost of a traditional phone screen, ₹85-150 per candidate, gives a similar answer: 400 screens would cost ₹34,000-60,000 to buy.

The screening hours most calculations leave out

Screening labour goes missing for three reasons.

Salaries are treated as fixed. The recruiting team’s pay is in the HR budget whether they make 20 hires or 200, so some teams leave it out of cost per hire altogether. That makes in-house hiring look nearly free and agency hiring look expensive.

Time is averaged, not measured. A recruiter’s salary is spread evenly across all their hires. A role that needed 400 phone screens and a role that needed 15 end up with the same internal cost per hire.

Hiring manager time is nobody’s line. It sits in another department’s budget. AIHR lists it as an internal cost; most dashboards still do not.

The result is a metric that rewards the wrong behaviour. Cutting a ₹75,000 job portal spend shows up immediately. Saving 80 recruiter hours on phone screens shows up nowhere, even though it was worth more in the example above. We wrote about where those hours actually go in the 18-hour phone screen.

Why the lowest cost per hire is not the goal

A cheap hire who leaves in three months costs more than an expensive one who stays two years. Cost per hire measures the spend to fill a seat, not whether the seat stays filled. That is why it should always be read next to a quality measure: 90-day retention, offer-to-joining rate or hiring manager satisfaction. We have argued before that the cost of a bad campus hire belongs on the P&L, and it belongs next to this number too.

There is also a speed side. A screen that takes 10 days to work through 400 candidates loses the best ones to faster employers. That cost never appears in cost per hire either, but it shows up in time to hire.

How to lower cost per hire without lowering the bar

1. Measure screening hours for one role. Before changing anything, time the resume reading and first-round calls for one hiring round. Most teams are surprised.
2. Cut applicants who were never a fit, earlier. Clear knockout criteria in the job post and the first screen stop hundreds of irrelevant applications from consuming recruiter time.
3. Stop phone-screening everyone who passes the resume stage. A recruiter call is the most expensive way to find out whether someone can explain their own experience.
4. Spend hiring manager time only on real finalists. Their 30 minutes is worth more than a recruiter’s.
5. Track source quality, not just source cost. A cheap portal that sends 1,000 irrelevant resumes is not cheap.

Where AI screening changes the number

The line AI screening moves is the internal one. In HireQwik, resumes are scored against the job automatically, and candidates above the cut book their own 10-20 minute AI voice interview at a time that suits them. The recruiter reviews verdicts with the transcript and evidence instead of making the calls. In our pilot, that turned about 18 hours of manual profile review into 1-2 hours of HR review for a 3,000-applicant drive, and the screening phase took 1-2 working days instead of 6-10.

In the worked example, if the 105 screening hours became 10 hours of review, internal cost would fall by about ₹47,500 for the round. Add the cost of the tool before you celebrate, which for us is well below a manual phone screen; book a demo and we will run your own role’s numbers with you. The point of the exercise is not our product. It is that once you count screening time honestly, you can finally see which part of hiring is expensive.