How much does it cost to onboard a new employee? Priced line by line

Ask how much it costs to onboard a new employee and the top-ranking answers quote three different figures, each attributed to SHRM: $4,129, $4,425 and $4,700. In the results we sampled (10 August 2026), none of the three carried a date, and the figure we could trace to a dated SHRM publication (April 2022) turns out to measure the cost of hiring, not onboarding.

So here is the honest version. For a small US business, the cash cost of onboarding one employee is often modest: some hardware, a few dollars of software, a training invoice if the role needs one. The real cost is paid time. Priced line by line at the US private-sector all-in employer cost at the median wage, $34.78 per hour worked (BLS, March 2026), our worked example below comes to about $7,756 for the first 90 days, and roughly 87% of it never appears on an invoice. This guide shows the four lines, the arithmetic, and where the number actually moves.

Why every onboarding cost figure you have read disagrees

The short answer: most published numbers measure hiring, count only cash, or carry no date, so they cannot agree with each other, and none of them can anchor your budget. Before you can price onboarding you have to draw the line between it and recruiting, and most of the articles ranking for this question do not.

The traceable figure is real, it just belongs to the neighbouring question. SHRM’s benchmarking research (April 2022) put the average cost per hire at nearly $4,700: the recruiting side of a hire, meaning job ads, recruiter time and interviews. The same article notes that many employers estimate the total cost to hire a new employee at three to four times the position’s salary, and quotes a talent economist’s split of those hiring costs: 30 to 40 percent hard cost, the other 60 percent soft cost, meaning people’s time. Onboarding sits outside that measurement: it is the phase that begins after the hire the $4,700 paid for. But the same pattern, people’s time dwarfing the invoices, is exactly what you find when you price it.

The unattributed numbers are less useful. Ranges like “$600 to $1,800 for small companies” circulate with no source and, judging by their size, count only cash outlays. That is not wrong so much as incomplete: it prices the laptop and ignores the month of half-speed work that costs four times as much.

The First-90-Days Ledger

Every cost of onboarding, from the first admin hour to the day-90 close, fits on four lines, and each line has a knowable source: your own hours, your own invoices, and one government statistic. I call it the First-90-Days Ledger because the window matches how onboarding actually runs: it should close, on purpose, around day 90.

Ledger line	What it contains	Where the number comes from
1. Admin before day one	Contract, tax forms, payroll setup, accounts, equipment ordering, scheduling	Your ops hours × a loaded hourly rate
2. Manager and buddy time	Day one, week-one sessions, weekly one-to-ones, 30-day and 90-day reviews	Your plan’s hours × the same rate
3. Hardware, software, training fees	Laptop, peripherals, the new seat in your tools, any third-party training or certification invoice	Actual invoices and published prices
4. The ramp gap	The share of the new hire’s paid hours not yet producing full output	Paid hours × ramp percentage × the same rate

Two boundaries keep the ledger honest. Recruiting spend (job ads, recruiter time, background checks) belongs to the hiring bill, not this one. And employer payroll taxes need no line of their own: they ride inside the loaded rate, because the BLS figure below counts legally required benefits as part of the hourly cost.

The loaded hourly rate is the one input people get wrong, usually by forgetting that an hour of employee time costs more than the wage. The Bureau of Labor Statistics measures exactly this: in March 2026, total employer cost for private-industry workers averaged $46.60 per hour worked, of which wages were 69.9% and benefits 30.1%. At the median wage, a private-industry worker cost $34.78 per hour all in; BLS publishes that total-compensation figure directly, by wage percentile. Two useful rules fall out of the release: multiply a mid-range wage by about 1.4 to load it (the load runs lighter at the low end of the wage scale), and if you want a defensible middle-of-the-road rate, $34.78 is the published all-in cost at the median.

Lines 1 and 2 come from whatever onboarding plan you run. Ours is public: a day-by-day onboarding plan built on a 1–7–30–90 spine, and the hours below are that plan, priced.

The worked example: one hire at a ten-person company

At the BLS median all-in rate ($34.78 an hour, March 2026), a full-time hire onboarded on our published plan comes to about $7,756 across the first 90 days in our worked example, and $5,913 of it is the ramp gap. Here is the complete ledger, with every working assumption visible so you can overwrite it for your own team.

The setting: a ten-person US company, one full-time hire at the median loaded rate of $34.78 an hour, 40-hour weeks, 13 weeks from day one to the 90-day close.

Ledger line	Basis	Cost
1. Admin before day one	5 hours × $34.78	$174
2. Manager and buddy, day 1 to day 90	18 hours × $34.78	$626
3a. Hardware	example invoice: one $1,000 laptop	$1,000
3b. Software, one quarter	Zoho People ×10 users + one Gusto payroll seat (no third-party training fees in this example)	$43
4. Ramp gap over 13 weeks	170 hours × $34.78	$5,913
Total, first 90 days		$7,756

Where the hours come from:

• Admin, 5 hours. Contract out and back, W-4 and state forms sent, I-9 prepared, the state new-hire report, adding the hire to payroll, creating accounts, ordering and testing equipment, scheduling the first two days. Our estimate for a practised run; your first time will take longer.
• Manager and buddy, 18 hours. Day one is about three manager-hours. The rest of week one (assigning the first task, checking it landed, the end-of-week one-to-one) is three more. Weekly one-to-ones, training oversight and the 30-day review add about four across the first month; fortnightly check-ins and the 90-day conversation add three across months two and three. A buddy fields questions for roughly five hours over the quarter.
• Software, $43 a quarter. The cheapest priced path from our onboarding software round-up for small teams: Zoho People’s Essential HR at US$0.84 per user per month billed annually (captured 8 August 2026) for ten users over the three months, plus the new hire’s $6 monthly seat on Gusto, which we run our own payroll on.
• Ramp gap, 170 hours. Call it 50% effective output in month one (160 paid hours, 80 of them still ramp), then 75% across months two and three (360 paid hours, 90 of them ramp), 80 + 90 = 170 ramp hours in total. Those two percentages are working assumptions for a mid-level role, not research findings: overwrite them with your own reading of the job. A senior hire may ramp slower on purpose; a till operator will be near full speed in days.

Two things stand out. The cash you can point to (lines 3a and 3b, about $1,040) is around 13% of the total; the rest, about 87%, is paid time. And the ramp gap alone is 76% of the bill. The pattern the SHRM economist described for hiring, soft costs outweighing hard ones, holds at least as strongly in this onboarding-only ledger.

How to calculate onboarding costs for your own hire

To calculate onboarding costs, price four lines: (admin hours + manager hours + buddy hours) × loaded rate, plus hardware, software and any training fees at invoice, plus paid hours × ramp percentage × the same loaded rate, counted to day 90. The loaded rate is the hourly wage × roughly 1.4, or $34.78, the published all-in cost at the US private-sector median wage.

Three swap-ins do most of the work:

• The rate. If your ops person, manager and new hire earn very different amounts, price each line at its own person’s loaded rate instead of one blended figure. The BLS release publishes total-compensation percentiles ($18.06 at the 10th, $89.70 at the 90th): use them directly as loaded rates if you want brackets.
• The hours. If you have never onboarded anyone, borrow ours above and correct them after your first hire. If your plan is heavier (regulated training, certifications), line 2 grows and so does the ramp window.
• The ramp. This is the input worth ten minutes of honest thought, because it is the biggest line. Anchor it to milestones you can check rather than a feeling: when did the last hire first ship real work, and when did you stop reviewing everything they did?

One warning on precision: the ledger’s job is not a to-the-dollar answer, it is to stop you managing the $1,000 line while ignoring the $5,900 one.

What software adds to the bill, and when it subtracts

On 2026 prices, onboarding software bought at the cheap end is the smallest line on the ledger: the lowest priced tier we captured runs under a dollar per user per month, against thousands of dollars of paid time. The segment stretches to a $250 monthly floor, though, which over a quarter would outweigh the manager line itself, so the claim belongs to the cheap end, and that spread is the entire buying argument in both directions.

The real prices, from our priced round-up and the HR software pricing index: Zoho People’s Essential HR tier at US$0.84 per user per month billed annually (8 August 2026), Homebase’s All-in-One tier at $96 to $120 a month per location with new-hire packets (8 August 2026), Gusto’s $49 plus $6 per person with self-onboarding included (tested; captures June 2026), BambooHR’s $250 monthly floor for 25 or fewer employees (8 August 2026), and Deel’s HR plan from $14 per worker per month for cross-border teams (July 2026).

The arithmetic cuts both ways. If a $25-a-quarter tool reliably takes two hours off line 1 per hire, it pays for itself from two hires a year (about $139 of admin time back against $100 of annual fees). But software does not shorten the ramp gap, which is three-quarters of the bill, and buying it before your process exists automates a process you do not have. If hiring is occasional, run the plan manually; when the pre-boarding sequence becomes routine work, wiring it up is its own project (how to automate employee onboarding), and where onboarding sits among everything else worth systematising is mapped in our guide to which HR processes can be automated. If the trigger for all this is your first employee, the payroll seat is the one tool you cannot skip: choose payroll software by your hardest constraint, or start from the genuinely free payroll options if the budget is zero.

Where the real money leaks

The expensive onboarding failures all attack the same line: they stretch the ramp gap. Only 12% of employees strongly agree their organization does a great job of onboarding new employees, per Gallup, and the gap between a run plan and an improvised one is measured in weeks of half-speed output at full pay.

The arithmetic makes the leaks concrete. A week of a new hire’s paid time lost to missing accounts and an absent manager is 40 hours at $34.78: $1,391 paid against near-zero output, more than the laptop. A first month without a real task or a review does the same quietly, which is why our plan’s pass/fail line (one small piece of real work shipped by Wednesday of week one) exists: it surfaces blocked access while the meter has barely started running.

The worst case restarts every meter at once. A hire who walks at month three because onboarding never really happened takes the hiring spend (SHRM’s benchmark average: $4,700) and this entire ledger with them, and both bills run again for the replacement. Retention arguments for onboarding usually arrive as percentages; on a ten-person team it is simpler than that. You paid $7,756 to get one person to full speed. Losing them is the only way to pay it twice.

Knowing how much it costs to onboard a new employee will not make the bill smaller on its own. But the ledger tells you where management effort pays: the admin line responds to preparation, the software line barely matters at small-team prices, and the ramp line, three-quarters of the total, responds to exactly one thing, which is running a real plan and closing it on a date. Price your next hire’s first 90 days on the four lines before they start, and you will know, in dollars, why the week before day one is the cheapest week to work hard.