People & Ops

The true cost of an employee is base pay plus four layers the offer letter never shows: the statutory 7.65% employer FICA share (IRS Topic 751), benefits, one-time recruiting and equipment costs, and manager hours. This page's worked $85,000 illustration totals $121,483 in year one, a 1.43 multiple over base.

The offer letter says $85,000. The seat costs more, and the extra arrives in layers with different clocks: some per paycheck, some per month, some once, and one, the manager's attention, on no ledger at all.

The statutory layer. The employer's federal payroll tax share is 7.65% of wages: 6.2% for Social Security and 1.45% for Medicare(IRS, Topic 751). On an $85,000 base that is $6,503 a year before anything optional, and it is the only number in this section that is law rather than choice.

Federal and state unemployment taxes ride on top, and the state portion varies with your state and your own claims history, which is why the calculator below treats the rate as editable instead of pretending one figure covers every reader.

The monthly layer. Benefits recur whether the month went well or not: the employer share of the health premium, the retirement match, the small subscriptions that attach to a seat. A $750 monthly load is $9,000 a year that never once appears in a comp conversation, which is exactly why it gets forgotten when owners quote what they pay people.

The one-time layer. Recruiting spend and the equipment the seat needs on day one land in the first quarter and then wash out. They belong in year-one math and in nothing after, which is why this page keeps two totals instead of one.

The manager layer. The least billed and the most real. Every hire consumes reviewing, correcting, and answering from whoever runs them, and that person's hours were already the scarcest thing in the building. Four hours a week against a manager whose own loaded cost is $60 an hour is $12,480 a year, invisible on every statement you produce, paid in full anyway.

Stack the layers on the $85,000 estimator and year one reads like this, as an illustration built for reruns with your own numbers, never a market statistic:

The stack totals $121,483 in year one, a 1.43 multiple over base, and about $100,503 a year ongoing once the one-time costs wash out and the manager load settles. Divide the ongoing figure by a 2,080 hour year and the seat costs $48.32 an hour. The salary alone divides to $40.87. If your quotes bill estimating time at the second number, the gap compounds quietly on every job you win, and the fix belongs in your pricing, which is the Growth Signals guide's territory.

Keep the two totals for two different jobs. The ongoing figure is the budgeting and quoting number: it is what the seat costs in any normal year, and the loaded hourly built from it is the floor under any billing rate the seat supports.

The year-one figure is the decision number: it is what a yes actually commits over the next twelve months, and it is the honest comparison against the revenue or relief the hire is supposed to buy. Owners who use the ongoing number to make the hiring decision understate the commitment by the whole first-quarter stack; owners who use the year-one number to set billing rates overprice themselves forever. Same seat, two totals, each with its own job.

And year one is tighter than the multiple suggests, because the hours are not all productive ones. A hire who runs four months at half output delivers about 1,733 productive hours in year one, not 2,080, which pushes the illustration's cost per productive hour to $70.09. That number, the ramp-adjusted one, is the honest rate the business pays for output while the hire learns your shop, and it is the number the calculator below exists to put in front of you before the offer goes out.

The paycheck is the entry fee.

A comp band you have not reviewed in two years is still being reviewed, just not by you. The chain runs: the wage print rises, the market around you reprices its postings, a recruiter calls your best technician with the market's number, and your band either answered that call in advance or answers it in a panic, under notice, at counteroffer prices.

The print doing the work is average hourly earnings: $36.47 to $37.62 in the twelve months to July 2026 (FRED, CES0500000003),a 3.15% drift computed straight from the level series. Read it as the tide under every band you maintain, with one caution: you hire a specific trade in a specific county, and the national series is the baseline, never the quote. When your region's skilled-trade postings stay hot while the national drift cools, your band answers to the local number.

Discipline upward means moving bands on a calendar instead of under duress. A raise granted at review time moves one salary and lands as recognition. The same dollars granted as a counteroffer, after the resignation letter, cost more twice over: the match itself usually runs above the planned raise, and the shop learns that notice is the negotiation channel, so the next outside offer gets solicited instead of fielded.

The practice owner's three-week front desk vacancy is usually the same lesson from the other side: when a posting runs dry that long, the band it advertises was set against a print two years stale, and the vacancy is the market's counteroffer.

Discipline downward exists too, and almost nobody runs it. When openings fall and quits flatten, as the current prints show, the switching premium shrinks, and the peak-market number you once paid to fill a seat does not have to become the permanent baseline for every seat after it.

New seats price at today's market. The symmetry is the discipline: if a hot print justified moving the band up, a cooled print earns the right to hold it flat, and the operator who cannot bring himself to do the second half has a ratchet, which is a cost structure, wearing a compensation philosophy.

Building the band takes an evening, and the materials are free. Pull the national drift from the earnings series above, then pull three live local postings for the same seat, which are the offers your people can actually walk into. Add your own last two offers for the role, accepted or declined, because a declined offer is a price signal you already paid for.

The band that falls out of those five data points will not be precise, and it does not need to be: its job is to be reviewed on a calendar and moved for reasons you can name, so that when the recruiter calls your technician, the number on file is one you chose recently rather than one you inherited from a different market.

One more mechanism keeps owners honest here: internal equity. A single matched outside offer quietly reprices the whole crew, because the others find out, and they always find out. At that point the choice is matching internally or staffing the resentment, and both cost money; the only cheap option was the band review that would have made the outside offer uncompetitive in the first place. Twenty minutes per band, twice a year, against the print and two local postings.

Your comp bands are being set in public.

That is what settled churn looks like: twenty-six months between 1.9 and 2.2 (FRED, JTSQUR). A flat, low quits line means the outside market is winning fewer of your people away, which lowers the temperature under every retention decision you own and makes this the cheap season for the habits in this section. The gauge is worth a monthly glance for exactly one reason: when it moves, the price of keeping people moves first, and the print is public a quarter before the pressure reaches your shop.

Then one person quits anyway, and the average stops mattering. The chain that costs money runs: the quit takes the unwritten knowledge, the unwritten knowledge was load-bearing, and the first place it fails is customer-facing. The contractor's estimator again: the margin rule for the county with the inspection quirks, the customer who must never be quoted from the standard sheet, the supplier whose list prices are an opening position.

None of it lives in the estimating software. Three weeks after the goodbye cake, a quote goes out priced from the file instead of the memory, and it is wrong in whichever direction hurts more: too high and the bid is lost, too low and the job is won at a loss that surfaces in ninety days.

Regretted attrition is the operator's term for the quits you would have paid to prevent, and its math needs no industry multiplier, because every piece is priceable from your own numbers: the vacancy months (covered by overtime, by you, or by nobody), the replacement's recruiting and ramp at the calculator's rates, the manager hours of retraining, and the error rate while the memory rebuilds.

Run the calculator on the replacement seat, add the vacancy, and put the total next to the raise that was on the table six months earlier. That comparison, run before the resignation instead of after, is the whole discipline.

When the quit lands anyway, the two-week notice period is an asset with a fast expiry, and most owners spend it on farewell logistics instead of extraction.

The triage that protects the quotes: the departing person and their manager walk every live job and every open bid together, out loud, with the successor in the room; the one-page seat file gets written or corrected while its author can still be asked questions; and the handful of customer and supplier relationships that ran on the person rather than the company get a joint call, so the relationship transfers instead of lapsing.

None of it takes more than a few working days, and none of it can be done three weeks later.

Three habits price against it. Write the tribal knowledge down while its owner still works for you: the pricing quirks, the customer rules, one page per seat, refreshed quarterly, owned by the seat's manager. Hold stay conversations on a calendar rather than after resignations, because the version held after is a counteroffer with worse manners.

And separate regretted from non-regretted in whatever turnover figure you track: a number that is all non-regretted describes a functioning filter, and the same number with your two best people inside it describes an emergency hiding in an average.

Institutional memory gives two weeks notice.

Nine inputs, all reachable tonight: the offer letter, the benefits invoice, the last recruiting spend, and an honest guess at the ramp. Five outputs that decide things: the fully-loaded cost of year one, the ongoing annual cost after it, the loaded hourly on a 2,080 hour year, the year-one cost per productive hour with the ramp priced in, and the multiple over base salary.

If you searched for an employee cost calculator or a true cost of an employee calculator, this is that instrument, with the ramp arithmetic most versions skip.

Run it three ways. The seat you are about to post tells you what the offer letter actually commits, which belongs in the quote rates that seat will produce. The replacement for the person you are worried about, plus your vacancy months, prices the quit before it happens, which is the regretted-attrition number from the section above. And the seat you filled last year, run with real figures, calibrates how honest your ramp guesses were, which makes every future run better.

Hire now when the seat's throughput is already being bought some other way: work you turn down that you could price profitably, overtime that stopped being seasonal, or your own evenings covering a job you should hand off. Wait when demand is one customer deep, or when you cannot write the seat's first ninety days on one page.

The case for waiting always sounds like prudence, and sometimes it is. The way to tell is to price both sides, and the true price of hiring hides in the ramp. A hire lands, the ramp begins, the manager's week loses hours to teaching, the crew's most experienced hands slow down to answer questions, and only then does the new capacity arrive.

Team throughput dips before it rises. An owner who prices the seat at salary and expects week-one output has mispriced both sides of the decision, which is how good hires get labeled mistakes in month two.

The demand test. Hire when the seat's throughput is already being bought some other way. Work you turned down that you could have priced profitably is the cleanest signal, because it comes with its own arithmetic: when the declined work's margin runs above the seat's loaded cost from the calculator, the market has already voted, and waiting is a standing order to keep declining it.

Overtime that stopped being seasonal is the same signal wearing payroll clothes: you are paying premium rates for straight-time work. And your own hours count too; the owner covering estimating nights is spending the most expensive labor in the company on a seat the calculator can price.

The durability test. Demand one customer deep is concentration risk dressed as growth, and a hiring decision made on it staffs the concentration. Look for the second and third source of the same demand before the requisition, or size the hire to the demand you keep if the big customer walks.

The readiness test. If you cannot write the seat's first ninety days of output on one page, the hire buys payroll while the plan gets written on company time. The page costs an evening; the ramp it shortens costs months.

The prints set the backdrop, never the decision. Openings at 7,359,000 and down two prints in a row, quits flat at 2.0%, unemployment at 4.1% (June and July 2026, FRED): for a hiring owner that mix reads easier than it did earlier in the year, with fewer postings competing against yours and candidates less likely to be countered away mid-process. What the backdrop cannot see is your pipeline, which is why the three tests come first and the prints come second, as a thumb on the scale for timing.

Between yes and wait there are smaller units of purchase, and they exist to test the demand before the full commitment. A part-time hire, a contractor, or overflow work sent to a peer shop each buys a fraction of the seat's throughput at a worse hourly rate but with no ramp, no benefits stack, and no severance conversation if the demand proves shallow.

The arithmetic is the same calculator run sideways: when the premium you pay for the flexible version, month after month, grows past the loaded cost of the real seat, the test has ended and returned an answer. Operators get this wrong in both directions: some hire full-time to relieve a two-month bulge, and some pay contractor premiums for two years to avoid an offer letter the demand justified in the first quarter.

Waiting has a price of its own, and it is the ramp shifted right: the seat you defer a quarter reaches full output a quarter later than that, because the ramp waits for the start date. If the demand test is already passing, deferral is turned-down revenue with a start date attached. If it is failing, deferral is free. That asymmetry is the whole hire-now-or-wait question compressed to one sentence, and it is answerable from your own numbers in an evening.

You are buying throughput, on a delay.

Obligations arrive on two clocks. The calendar clock is visible: policy renewals, the workers compensation audit, filing deadlines, the annual notices your payroll provider nags you about. The headcount clock is the one that catches operators, because federal and state employment mandates step in at published employee-count thresholds, and crossing one mid-year comes with no letter in the mail. The count moves every time you hire; the obligations attach whether you noticed or not.

The mechanism worth respecting is the silence. A threshold crossed is an obligation acquired at the moment of the offer letter and discovered at the moment of an audit, a claim, or a former employee's lawyer. The distance between those two moments is where the expensive versions of compliance live, and it is also entirely yours to close, because you are the only party who knows a hire is coming before it happens.

The countdown habit is one page and a recurring half hour. On the page: every obligation you carry, its owner, its date, and its lead time; below that, every threshold you are approaching, with your current headcount written next to it. Review the page when you close the books each month and again before every offer letter goes out.

Put the dates in the same calendar that runs payroll, because that is the calendar that actually gets read. Which thresholds bind at which counts in your state is a question your payroll provider, your benefits broker, and your accountant each answer for free; the habit's whole value is asking them the quarter before the crossing instead of the quarter after.

The countdown also feeds the hiring math. A marginal hire that crosses a threshold carries the threshold's cost with it, at whatever your broker and accountant quote for the new obligations, and that belongs in the fully-loaded picture before the offer goes out.

The calculator holds the seat's own costs; the countdown page holds the costs the seat triggers. Read together, they are the honest price of growing by one. Financing that growth, and what your bank's covenants say about a heavier payroll, is the Capital for Operators guide's side of the map.

Deadlines found early are paperwork. Found late, they are findings.