Employer guideUpdated September 2026

Hiring Metrics Benchmarks: The Two We Measured, and the Four We Will Not Fake

Across 891 campaigns and 1,334 campaign-months in 2026 the median campaign delivered an applicant for $13.88, the volume-weighted average was $8.02, and the middle half of campaigns ran between $6.48 and $29.74. The metric that explains the gap is not the ad auction. Application completion rate explains 70% of the variation between campaigns; what impressions cost explains 18%. Those two numbers we measured. Time to fill, cost per hire, quality of hire and offer acceptance rate we did not, because they live in your applicant tracking system and no advertising platform can see them.

Cost per applicant is a distribution, not a number

The single most common error in a job advertising budget review is quoting one figure. There are two defensible averages in this dataset and they are nearly double each other: the volume-weighted average of $8.02 and the median campaign at $13.88. Both are correct. The average is pulled down by a handful of very high-volume frontline roles that convert exceptionally well, so it is the right number for sizing an annual portfolio and the wrong number for budgeting one requisition.

Cost per applicant by role family

Fifteen role families, with the median and the range the middle half of campaigns actually landed in. Read the range rather than the median; a single number invites the wrong argument with a finance team. Apply rate is the share of people who clicked and then completed the form, and it is the column that explains most of the rest of the table.

Boostpoint 2026 Social Job Advertising Benchmark: 891 Boostpoint-managed campaigns on Meta, 1,334 campaign-months, 2026. Sorted by median. Costs are what advertisers paid for advertising only.

The pattern that runs down the table is not what most people expect. Expensive roles are not expensive because attention costs more. Driver campaigns bought impressions more cheaply than caregiver campaigns, at $17.88 per thousand against $27.88, and still cost roughly six times more per applicant. The whole difference is the apply rate: 8% of drivers who click finish a form, against 31% of caregivers. You are not paying for reach. You are paying for friction.

The benchmark that actually moves the number

We tested how much of the variation between campaigns each input explains on its own. Application completion rate explains 70%. Click-through rate explains 30%. What the auction charges per thousand impressions explains 18%. Nearly every hour a team spends negotiating media efficiency is aimed at the smallest of the three levers.

41% of all budget ran in campaigns where fewer than one in ten clickers finished the form, and those campaigns returned 11% of all applicants. Campaigns above 35% used 6% of budget to deliver 28% of applicants, at $1.61 each. That is a twelvefold cost difference between the bottom band and the 20 to 35% band, and it is mostly structural: whether the application stays on the platform, how many questions the form asks, and whether the screening removes people you could never hire or people you would have hired after a five-minute conversation.

Recruitment KPIs: what each one is, and where its number actually lives

“Recruitment KPIs” and “talent acquisition metrics” are the same list under two vocabularies — the second is what large employers call it. What matters is not the naming but which system holds each number, because that determines whether you can trust it and whether anybody outside your business could benchmark it for you.

Metrics your advertising can see. Cost per applicant, application completion rate, impressions, frequency and cost per thousand. These are measured directly, which is why they are the ones published as distributions further up this page.

Metrics only your ATS can see. Applicants per hire, screening pass rate, interview-to-offer ratio, time to fill and time to hire. Nobody outside your business can measure these, and an external benchmark for them is somebody's survey rather than a measurement. Where those surveys do exist, we set them out with their sources on average cost per hire by industry.

Metrics only your HR system can see. Cost per hire, offer acceptance rate, quality of hire, ninety-day retention and first-year turnover.

Offer acceptance rate is the simplest of these and the most under-used. It is offers accepted divided by offers extended, over a period. A rate below about four in five is usually telling you one of three things — the pay in the advert did not match the pay in the offer, the process took long enough for a competing offer to land, or the job as described at interview was not the job as described in the offer letter. It is worth calculating monthly because it moves fast and it is the earliest signal that something upstream has broken.

Quality of hire is the metric everyone wants and almost nobody defines consistently. There is no single formula. In practice employers build it from some combination of performance rating at six or twelve months, retention past a threshold, hiring manager satisfaction, and time to full productivity — weighted and rolled into an index. Because the inputs are chosen by each employer, a published quality-of-hire benchmark compares nothing to nothing. The version that is useful is internal and relative: score your hires the same way every quarter and watch whether the number moves, by source, by role and by hiring manager.

That is the practical rule for this whole list. Advertising metrics are worth benchmarking externally because they are measured the same way everywhere. Everything downstream of the application is worth tracking rigorously and comparing only against your own last quarter.

Recruiting metrics to track: the formula and the benchmark for each

The section above sorts recruiting metrics by the system that holds them. This is the working list: twelve metrics a frontline employer can track every month, how to calculate each one, and the benchmark on this page to hold it against. Where the benchmark column says “your own last quarter,” that is deliberate; the number lives in your records and nobody outside can measure it for you.

Two notes on using the list. Track the first four weekly per campaign, because they move within days and are the cheapest to fix; track the rest monthly by role and by source, because a single month of hires is too few to read. And put the market beside them: in the BLS JOLTS quits table, the July 2026 (preliminary) quits rate was 1.9% for all nonfarm employers and 3.4% in leisure and hospitality, so a hospitality operator with a quit rate near that level is running with the market, not behind it. The nine-column hiring tracker further down this page puts most of these into one sheet.

The four metrics we are not going to publish a number for

Search for hiring metrics benchmarks and you will find confident figures for time to fill, cost per hire, quality of hire and offer acceptance rate. Some come from employer surveys with a few hundred self-selected respondents. Some are extrapolations from advertising data. A submitted form is not a screened candidate, an interview or a hire, and no advertising platform can see what happened after your recruiter picked it up. So the honest table is the one that says which side of that line each metric falls on. If what you need is the cost of the days a seat stays open rather than a benchmark for them, that is the cost of vacancy calculator.

Our own classification of what our data can and cannot support. We would rather leave four rows blank than publish a figure we would have to caveat into meaninglessness.

The one public benchmark you can check yourself

If you want a hiring benchmark that is neither ours nor a vendor survey, the Bureau of Labor Statistics publishes one every month and it is free. In July 2026 the national hires rate was 3.2% and the total separations rate was also 3.2%. That is the whole story of the hiring market in two numbers: at the national level, hiring is running at exactly replacement. The quits rate was 1.9% overall, 3.4% in leisure and hospitality and 3.1% in retail, which is why a frontline operator experiences a labor market that the headline figure describes as calm.

Frequency: the early warning nobody watches

Frequency is the average number of times each person in your targeted audience saw your ad in a month, and it is the most reliable leading indicator in the dataset. Below roughly 2.5 the median cost per applicant sits between $11.48 and $13.39. Above 3.0 it rises to $18.76, and above 4.0 to $26.78, while click-through rate falls from 1.69% to 0.94%. The audience has seen it and stopped responding. 33% of all budget in the study ran above a frequency of 3.0, which makes it the clearest single pool of recoverable money we found.

What a normal campaign looks like, and what we still cannot tell you

Half of all campaign-months in the study ran on less than $334, returning a median of 20 applicants and reaching a median of 6,061 people. The upper quartile begins at $554 and the top decile at $988. If somebody has told you that social recruitment advertising needs a large committed budget before it works, that distribution is the counter-evidence, and it is drawn from delivery rather than from a pitch.

What we cannot tell you is the part that matters most to a hiring plan: what an applicant becomes. Applicants to screens to interviews to hires, by source, is the one series no advertising platform can produce, and without it every efficiency decision is made half-blind. If you change one thing this quarter, measure apply rate per campaign and treat anything under 10% as a defect to investigate rather than a cost to absorb. The full distribution, the methodology and a six-question worksheet are in our 2026 social job advertising benchmark.

These metrics are the measurement half of a plan. The annual document they feed — the demand forecast, the channel mix priced per applicant, and the owner against each line — is on talent acquisition strategy.

Application-to-interview and application-to-hire ratios

The application-to-interview ratio is completed applications divided by interviews held; the application-to-hire ratio is completed applications divided by starts. They are the two numbers that turn a cost per applicant into a budget, and like every metric after the application they live in your applicant tracking system, so we do not publish a benchmark for either. What we can show is how they work together.

• They chain. Application-to-hire is application-to-interview multiplied by interview-to-hire. If you interview one applicant in three and hire one interviewee in four, you need twelve applicants per hire. The interview-to-hire half usually sits around 3:1 to 5:1 when the phone screen is doing its job, as the FAQ below sets out.
• They set your advertising cost per hire. Divide your cost per applicant by your application-to-hire rate. At the $13.88 median from the 2026 benchmark and one hire per ten applicants, that is $138.80 of advertising per hire, before recruiter time or anything else.
• They move with speed, not only with quality. An application-to-interview ratio that gets worse usually means slow first contact or no-shows rather than weaker applicants, which is why the hiring tracker below counts applicants contacted within one hour and interviews attended.

Staffing KPIs: the metrics that tell you whether the schedule is covered

Recruiting metrics describe the pipeline. Staffing metrics describe the result on the floor: whether the shifts you scheduled were worked by people you employ, and at what cost. A frontline operation needs both, because a pipeline can look healthy while the schedule is held together by overtime and agency staff. These are the staffing KPIs worth a line on a monthly report, with the calculation for each.

Read them as a chain rather than seven separate numbers. Turnover sets the vacancies, time to fill sets how long each vacancy stays open, and shift fill rate and overtime share show what those open days cost. None of these has an external benchmark worth quoting at the level of your own operation; the national context is the BLS JOLTS hires, separations and quits rates in the public benchmark section above, and our employee turnover rates by industry page sets it out by sector. The one advertising number that feeds this chain is cost per applicant, because it prices how quickly you can refill a vacancy; the role-family ranges from the Boostpoint 2026 Social Job Advertising Benchmark are in the table at the top of this page.

Recruiting metrics examples: one sample month, worked through

The formulas above are easier to use once you have seen them run on one set of numbers. The month below is an illustrative sample, not a benchmark and not a real client: one role, one site, one month of advertising, with every metric calculated from the same seven inputs.

The sample was built so the cost per applicant lands on the $13.88 median campaign in the 2026 benchmark, which makes the last line match the $138.80 example earlier on this page. Swap in your own seven inputs and every row recalculates the same way. The point of working a sample through is the order it reveals: the apply rate and cost per applicant are set before anyone in your business touches a candidate, while every row from attendance onward is set by your own process, and only your applicant tracking system can tell you whether those rows are improving.

The interview-to-hire ratio in reverse: how many applicants you need each week

The ratios above are usually read forward, from applicants to hires. Run them backward and they become a weekly plan: start from the hires the schedule needs, divide back through the interview-to-hire ratio, the interview show rate and the share of applicants you book, and you get the number of applicants the advertising has to produce each week. The worked example below uses the same illustrative ratios as the sample month above, so it is a method, not a benchmark. For the planning question of how many people to interview per opening, see how many candidates to interview.

The $13.88 is the median campaign in the Boostpoint 2026 Social Job Advertising Benchmark; replace it with your role family's figure from the table at the top of this page, and replace every ratio with your own from the last quarter. Two things fall out of doing it this way. First, the interview show rate is as powerful a lever as the budget: lift it from 80% to 100% in the example and the same two hires need 16 applicants instead of 20. Second, the weekly number is what a campaign should be judged against, because a campaign that produces 20 applicants in a week the plan needed 20 is working, whatever its cost per applicant looks like in isolation.

Interview show rate is the same measure as the attendance rate in the recruiting metrics table: interviews attended divided by interviews scheduled. The levers that move it are on the applicant's side of the calendar: booking the interview in the first conversation rather than by email days later, offering slots within a day or two, and a confirmation and reminder by text. Our interview scheduling guide covers the booking side, and the no call, no show policy page covers what to do when a new hire does not turn up on day one.

Frequently asked questions

A hiring tracker in nine columns

Most operations that ask for a hiring tracker are handed either an ATS report nobody reads or a spreadsheet with forty columns that stops being updated in week three. Nine columns is the version that survives, because every one of them is either a date somebody already knows or a number that changes a decision.

Three ratios fall out of those nine columns without any further work, and between them they diagnose most hiring problems:

• Applicants per hire. Applicants divided by starts. This is the number that drives every budget on a hiring plan, and most employers are guessing at it.
• Attendance rate. Interviews attended divided by interviews scheduled. If this is below about two-thirds, the problem is not your applicant volume and no amount of extra advertising will fix it.
• Interview-to-hire ratio. Interviews attended divided by hires. Read it alongside the attendance rate, because the two fail for opposite reasons.

On that last one: the interview-to-hire ratio is a screening measure, not a market measure. A ratio of 2:1 means almost everyone you interview is hireable, which usually means you are screening hard on the phone — efficient, unless you are also screening out people who would have worked. A ratio of 8:1 means the interview is doing the screening the phone should have done, which is expensive in manager hours and in candidate goodwill. Somewhere around 3:1 to 5:1 is where most frontline operations sit when the phone screen is doing its job, and if yours is far outside that, the fix is in the screening stage rather than in the advertising.

Keep the tracker at one row per vacancy and update it weekly. The temptation is to add columns for source, referral, recruiter and reason-for-decline; resist it until the nine are being filled in reliably, because a tracker that is 60% complete tells you less than one with fewer columns and no gaps.