HR Statistics: Human Resources Data for Small Business

The first time I went looking for HR statistics, I was trying to work out whether losing three people in one year meant my company was broken or meant nothing at all. I found a great many pages of numbers and almost none of them helped, because almost none of them said which survey they came from, which year they described, or which kind of employer they described.

That is the standing problem with this category. A figure gets published once, quoted twice, and by the fourth repetition it has lost its sample size, its date, and its methodology. It still looks like evidence. It will still get pasted into a plan.

So every number below is traced to a primary source and labeled with its vintage: the Bureau of Labor Statistics for hiring, separations, and labor cost, Gallup for engagement and the behavior around quitting, and SHRM for the HR function itself. Where a popular statistic has no locatable origin, I say so instead of repeating it. And I read each figure the way a business without an HR department has to read it, which is the reason I built FirstHR and the reason most of this page is about denominators rather than headlines.

The Ten Numbers I Come Back To

These ten figures describe the current state of employment in the United States, each traceable to a named survey with a published date. If you read nothing else on this page, these are the ones that carry the most weight per word.

Two of them do most of the work in the rest of this guide. The quits contrast between industries tells you what normal churn looks like where you actually operate, which is the only benchmark that means anything. The 52 percent figure tells you how much of your own turnover is a management problem rather than a market one, and it is the number that changed how I ran my first team.

Turnover and Retention Statistics

US employers made 63.0 million hires in 2025 and recorded 62.8 million total separations, of which 38.0 million were quits and 21.2 million were layoffs and discharges. Hiring and leaving are close to the same size at national scale, which is why the labor market can feel frozen and frantic at once.

Per the Bureau of Labor Statistics Job Openings and Labor Turnover Survey release carrying the 2025 annual figures, the 2025 annual average rates were 3.3 percent for hires, 3.3 percent for total separations, 2.0 percent for quits, and 1.1 percent for layoffs and discharges. Those are monthly rates averaged across the year, not annual ones. Quits accounted for 60.6 percent of everyone who left a job, which is the single most useful ratio in the whole survey: most departures in the United States are still decisions employees make, not decisions employers make.

The all-industry figure is close to useless for benchmarking, though. Churn is an industry characteristic before it is a management characteristic, and the spread is enormous.

The annual column is an approximation produced by multiplying the monthly rate by twelve, and it drifts a little because the same seat can turn over more than once in a year. It is close enough to be useful and far more intuitive than a monthly rate. Source for both columns is the JOLTS annual average quits rates by industry table for 2025.

Compare your own number against your industry row rather than the national line. A restaurant with 30 percent annual turnover is doing considerably better than its sector; a manufacturer with the same 30 percent has a real problem. Both would look identical against the all-industry average, and that is exactly how people talk themselves into the wrong conclusion.

Then there is the cost side. Gallup puts the cost of replacing one employee at one-half to two times that person's annual salary and calls the range conservative, and estimates the total bill for voluntary turnover across US businesses at around a trillion dollars a year. The number underneath that headline is the one small employers should act on: 52 percent of people who left voluntarily said their manager or organization could have done something to prevent it, and 51 percent said no one had spoken to them about their job satisfaction or their future in the three months before they resigned.

Employee Engagement Statistics

Thirty-one percent of US employees were engaged at work in the first half of 2026, unchanged from 2025, and 18 percent were actively disengaged. Engagement peaked at 36 percent in 2020 and has been flat at 31 percent since 2024.

Gallup's 2026 measurement draws on 43,262 responses from employed US adults collected in February and May. The component I would look at first is role clarity: 49 percent strongly agree that they know what is expected of them at work, up two percentage points, but still far below the 61 percent recorded in 2015. Barely half the working population is clear enough about the job to pick the strongest answer on offer. That is a management artifact, not a labor market condition, and it is fixable in a one to one.

The reason engagement is worth measuring at all is that it correlates with outcomes you can bank. Gallup's meta-analysis of 456 studies across 276 organizations, covering 112,312 work units, compares business units in the top quartile of engagement against those in the bottom quartile.

Read that table as a comparison between groups rather than as a set of levers, per the published Gallup meta-analysis. Moving your team into the top quartile does not hand you 23 percent more profit; the units that sit there differ in many other ways too. What the research supports is that the gap between engaged and disengaged teams is large and consistent across industries and countries.

What It Costs to Employ Someone

Employing a private industry worker cost $46.60 per hour worked in March 2026, split between $32.60 in wages and salaries and $14.01 in benefits. Benefits are 30.1 percent of total compensation, and most owners underestimate that share badly.

Per the Bureau of Labor Statistics Employer Costs for Employee Compensation release for March 2026, wages accounted for 69.9 percent of employer costs and benefits for the remaining 30.1 percent. For state and local government workers the benefit share is much higher at 38.5 percent of a $66.41 hourly total, which is why public sector comparisons quoted without context make private employers think their own benefits are stingy.

The number to carry around is the multiplier rather than the dollars. At the national average, budgeting a role by salary alone understates the true cost by more than 40 percent. If you are sizing a hire against revenue, work from fully loaded labor cost rather than the offer number, and treat the benefits line as part of total compensation rather than as overhead.

The composition of that 30.1 percent matters as much as its size, because access to each benefit varies sharply with employer size and small employers face worse terms on the expensive ones.

HR Department and Staffing Statistics

The median organization runs 1.98 HR staff per 100 employees, up from 1.58 in 2017. HR has been getting more expensive and more staffed at the same time, which is a useful backdrop for anyone deciding whether to hire their first HR person.

Applied literally, the median ratio says a twenty-person company needs four tenths of an HR person. That is precisely the position most small employers occupy: the work is real, it has deadlines attached, and it does not add up to a role. The Occupational Outlook Handbook puts the median HR specialist salary at $75,940, which is the number that decides the question for most owners long before any ratio does.

The HR expense line is the trend I would watch. Doubling from 1.2 to 2.4 percent of operating expense in eight years says the function grew faster than the businesses around it. For a small employer the practical response is not to skip HR but to keep it in systems rather than in headcount, which is what an HRIS is for and where a small business HR setup earns its keep.

Hiring and Recruiting Statistics

The median time to fill a nonexecutive role is 39 calendar days, and the average cost per hire is close to $4,700. Both are recruiting-side figures only, and neither includes the months of reduced output after someone starts.

The external hire rate is the row I find most interesting and the one that rarely gets quoted. Ninety-seven percent of nonexecutive roles are being filled from outside, up four points in a year, which means internal promotion has become rarer even as employers complain about the cost of recruiting. For a small business that is an opening rather than a warning: promoting from within costs a fraction of those 39 days and that $4,700, and it is one of the few advantages a small team holds over a large one, because you already know how the person works.

Bear in mind what the cost-per-hire figure excludes. It counts recruiting spend, not the ramp.

HR Technology and AI Adoption

Thirty-nine percent of organizations have implemented AI somewhere in the HR function, with 7 percent planning to this year and 31 percent reporting no plans at all. Adoption in HR trails the business overall, where 62 percent report using AI somewhere.

Per SHRM research covering more than 1,900 HR professionals, the measurement gap is wider than the adoption gap. More than half of HR functions using AI have no formal way of telling whether it worked, and only 16 percent use return on investment as a metric. Buying is running ahead of evaluating, which is how a category ends up with an expensive stack nobody can defend.

For a business without an HR department, the sequencing implied by that data is worth following. Fix the record first, automate the repeatable parts second, and add anything genuinely intelligent third. Our guides to HR automation and AI in HR cover what that looks like in practice, including which tasks are worth handing over and which ones create liability.

How These Numbers Change for a Small Employer

Every statistic on this page was produced by averaging over employers who are mostly much larger than you, and three specific distortions follow from that. Knowing which one is operating tells you whether to act on a figure or ignore it.

The first is the denominator problem. At fourteen employees, one resignation is a 7 percent turnover rate and two is 14 percent. A national rate computed over millions of workers is stable; yours moves in jumps large enough to look like a crisis or a triumph depending on the month. Report the raw counts next to any percentage you calculate, use a rolling twelve-month window rather than a quarter, and refuse to react to a single data point.

The second is that establishment is not company. Federal surveys often report by establishment, meaning one physical location. A twenty-person independent business and a twenty-person branch of a national chain land in the same bucket while having nothing in common in how HR actually gets done.

The third is that ratios assume a function you do not have. An HR staffing ratio of 1.98 per 100 presumes HR is a department with a budget line. In a small company it is a set of tasks distributed across an owner, an office manager, and whoever happens to be free, which is why the ratio should be read as a cost signal rather than as a hiring instruction.

The one national figure I would apply directly, without adjustment, is the 52 percent preventability finding. Small teams have shorter distances between an owner and a person thinking about leaving, which means the conversation that changes the outcome is easier to have and harder to excuse skipping. That advantage does not appear in any benchmark, and it is the strongest one you hold.

Statistics That Circulate Without a Source

Four HR figures appear on ranking pages constantly without a traceable primary source or with a badly out-of-date one. I am flagging them rather than repeating them, because an unsourced number that justifies a purchase is how a rumor becomes a budget.

The test I apply to anything, this page included, is whether you can name the survey, the sample, and the date. Every figure above passes it. If a statistic cannot, it may still be true, but it is not evidence, and it should not be the reason a small business spends money it has to earn back.

Four Ways These Numbers Get Misread

Most misuse of HR statistics comes from four specific errors, each of which I have made. They are worth knowing by name because they are easy to spot once you have.

Monthly rates get read as annual ones. A 2.0 percent quits rate sounds trivially small until you notice it is a monthly figure and the annual equivalent is roughly a quarter of the workforce. Whenever a labor turnover figure looks implausibly low, check the period before you celebrate.

Relative differences get read as absolute ones. A 23 percent difference in profitability between quartiles is not 23 percentage points of margin, and a study reporting that engaged teams have 43 percent lower turnover is not promising you a 43 point reduction. Both describe gaps between groups of companies that differ in many other ways as well.

Top-box scores get read as failure rates. When Gallup reports that 49 percent strongly agree they know what is expected of them, the other 51 percent are not all lost and confused; many of them chose "agree" rather than the strongest option available. The measure is deliberately hard, which makes it a good tracker over time and a bad headline.

Averages get read as descriptions. A blended cost per hire mixes executive searches with hourly roles, and a national quits rate mixes restaurants with banks. The average describes a distribution that may contain no company resembling yours, which is the entire reason the industry table above exists.

Your Own Versions of These Numbers

Every statistic on this page has a version you can calculate from data you already have: hire dates, termination dates, payroll, and a calendar. Six questions convert the national picture into a decision about your own business.

Answering those in your head is not the same as answering them on paper. The worksheet below puts the figures from this guide in one column and leaves the next one blank for yours, so the gap becomes something you can look at. The second sheet is for the statistics you meet everywhere else: a place to write down the survey, the sample, and the publication date behind a number before you let it into a budget conversation, with the four tests from this page at the bottom.

None of this needs an HR department. It needs the hire and termination dates you already have, a denominator you have agreed on in advance, and the discipline to redo it on the same schedule every year. Keeping those records in one place rather than across three spreadsheets is most of the battle, and it is the part FirstHR was built to take off your desk so that onboarding, documents, and the employee record all update from the same source.

Frequently Asked Questions