Three ways to fill a role, three different bills
Contingent labor is often the largest off-payroll spend an organization carries, and it usually sits outside the HRIS headcount where nobody prices it. The starting point is that the same role can be filled three ways, and each route bills differently. The comparison only works when all three are priced on the same wage base and the same period.
Agency temp Bill rate
You pay the agency bill rate: the worker pay rate times one plus the markup, with overtime billed at a premium. The agency runs payroll, remits the employment taxes, carries workers compensation, and screens the worker. You are buying flexibility and speed, and the meter runs for as long as the assignment does.
Temp-to-hire Trial, then convert
A trial window on the bill rate, then conversion to your payroll. The agency agreement usually charges a conversion or buyout fee, commonly 10% to 25% of first-year salary in 2024 to 2025 vendor surveys, and usually waives it after a threshold of worked hours that commonly starts near 520 and can run past 1,000. The waiver schedule is the lever most buyers never read.
Direct hire Salary plus burden
Salary plus your full labor burden from day one. If an outside recruiter fills the role, add a direct placement fee of 15% to 25% of first-year salary, with 20% the common benchmark. Highest commitment, lowest ongoing rate.
None of these is always the cheapest. Short remaining need favors the agency, long need favors payroll, and the buyout schedule decides the middle. A model that always recommends one route is a brochure, not a model.
What a 50% markup pays for
A 50% markup does not mean the agency keeps half the bill. The markup funds the employer obligations you would otherwise carry yourself, then the cost of running a staffing operation, and only then margin. Reading the markup as pure profit produces a savings claim that a CFO will take apart in one meeting, so the decomposition matters more than the headline number.
This decomposition is also why the markup range is so wide. Light industrial and clerical placements at volume sit near the bottom of the 25% to 75% band; specialized, low-volume, or high-risk placements sit near the top. Your agreement is the number that matters, and a markup transparency clause is the way to see it.
Build the burden number line by line
The conversion math turns on your labor burden, and a burden number nobody can defend forfeits the whole analysis. Two of the five lines are fixed by federal law, one is a federal tax that nets down to a small flat cost, and two are yours alone: your state unemployment rate and your workers compensation class rate. Build the number from the lines rather than borrowing an average.
How to read it. The 18% no-benefits default in our tools is a planning figure covering the statutory lines at a mid-range state rate; the 30% with-benefits default tracks the BLS benefits share. Both ship as editable inputs because SUTA and workers compensation are yours, not an average.
The four-way framework, on a worked example
The defensible way to price a conversion is four options side by side on the same wage base: stay on the agency bill, move the workers to a payrolling arrangement at a lower markup, bring them onto internal payroll without benefits, or bring them on with benefits. Take Calder Supply Company, a fictional distributor with 40 contingent workers at $20 per hour, working 40 hours plus 5 overtime hours a week for a remaining need of 26 weeks. The wage base, with overtime at time and a half, is $988,000.
The honest caveats belong in the model, not in a footnote. Conversion adds recruiting and ramp effort, a severance reserve on the exit path, and any buyout the agency agreement charges for taking the workers. When the remaining need is short, the buyout is high, or your loaded burden approaches the markup, the agency route wins, and a credible model shows that crossover instead of assuming conversion always pays.
Co-employment and misclassification, the hidden bill
Two legal exposures ride on every contingent program, and both concentrate in exactly the arrangements that look cheapest. The first is co-employment: long-tenure contingent workers who look and work like employees can claim employee benefits, and the anchor case is Vizcaino v. Microsoft, where a class of long-serving temps and contractors won the right to participate in employee benefit plans and the matter settled for roughly $97 million. The common corporate reflex, an 18-to-24 month tenure cap with a break in service, is exactly that, a common practice; it is not a legal safe harbor, and no tenure clock substitutes for getting the classification right.
The second is worker misclassification, and its federal status has three layers as of 23 July 2026. The Department of Labor 2024 final rule, a six-factor economic reality test effective 11 March 2024, remains on the books and still governs private FLSA litigation. Field Assistance Bulletin 2025-1 says the Department is not applying that rule in its own enforcement while it reconsiders. And a 2026 proposed rule, published 27 February 2026 with the comment period closed 28 April 2026, would rescind the 2024 rule but is not final. The IRS common-law test and state tests such as the California ABC test are independent of all of that, and state law can be stricter than federal.
California prices willful misclassification at $5,000 to $25,000 per violation. Labor Code section 226.8 sets civil penalties of $5,000 to $15,000 per violation, rising to $10,000 to $25,000 for a pattern or practice. If a conversion plan involves engaging former temps as independent contractors rather than employees, run the classification tests first and have employment counsel review the arrangement. The cheapest looking route on the spreadsheet can carry the largest bill on this page.
Where to run the math
The free Staffing Agency vs Payroll Cost Calculator runs the four-way comparison above in your browser on your own crew, rates, and markup, and the free Temp-to-Hire Breakeven Calculator prices the three paths for a single role and reads the buyout against the waiver schedule. Both use the sourced defaults on this page and leave every rate editable.
The paid line carries the same models further. The Contingent Labor Cost and Conversion Workbook adds the program-level rollup, a burden builder, and scenario compare; the Temp-to-Hire and Direct-Placement Decision Workbook models a whole requisition portfolio with a declining buyout schedule; the Contingent Workforce Headcount and Spend Tracker keeps the program visible month to month with a tenure-cap flag and a conversion watchlist; the Independent Contractor Classification and Risk Kit walks the three classification tests worker by worker; the Staffing Agency Management Kit fixes the agreement terms and scores the vendor against the markup benchmarks; the flagship Contingent Workforce Cost and Compliance Kit sequences the cost and compliance layers in one kit; and the Workforce Cost Control Bundle pairs the conversion workbook with the labor-cost planners for the wider what-does-this-labor-cost question. For the full employment cost picture behind the burden lines, see the fully loaded cost of an employee.
Where these figures come from
Primary sources
- Bureau of Labor Statistics, Employer Costs for Employee Compensation, March 2026 release. The source for total compensation averaging $46.60 per hour for private industry workers, benefits at 30.1% of total compensation, and legally required benefits at 7.2% of total compensation. bls.gov, ECEC latest numbersChecked 23 July 2026
- Internal Revenue Service, Topic No. 751, Social Security and Medicare withholding rates. The source for the employer FICA share of 7.65%: 6.2% Social Security up to the annual wage base, $184,500 for 2026, plus 1.45% Medicare with no wage cap, and for the additional 0.9% Medicare tax above $200,000 being withheld from employees with no employer match. irs.gov, Topic 751Checked 23 July 2026
- Internal Revenue Service, Topic No. 759, Federal Unemployment Tax. The source for the FUTA rate of 6.0% on the first $7,000 of wages and the 5.4% credit that brings the effective rate to 0.6%, about $42 per employee per year, outside credit-reduction states. irs.gov, Topic 759Checked 23 July 2026
- U.S. Department of Labor, Wage and Hour Division, independent contractor status under the FLSA. The source for the three-layer federal status: the 2024 final rule and its six-factor economic reality test, Field Assistance Bulletin 2025-1 stating the Department is not applying the 2024 rule in its own enforcement, and the 2026 proposed rule, published 27 February 2026, that would rescind it and was not final as of the check date. dol.gov, worker misclassificationChecked 23 July 2026
- California Labor Code section 226.8. The statute setting civil penalties for willful misclassification of independent contractors: $5,000 to $15,000 per violation, and $10,000 to $25,000 per violation for a pattern or practice. leginfo.legislature.ca.gov, section 226.8Checked 23 July 2026
- Vizcaino v. Microsoft Corp., 120 F.3d 1006 (9th Cir. 1997). The co-employment anchor case: long-tenure contingent workers held entitled to participate in employee benefit plans, with the litigation ultimately settling for roughly $97 million. The reason tenure caps became a corporate reflex, and the reminder that they are practice, not safe harbor. law.justia.com, Vizcaino v. MicrosoftChecked 23 July 2026
- Industry survey ranges for markup and fees, 2024 to 2025. Agency markup ranges of 25% to 75% of the pay rate (altLINE and Upwork 2025 survey data); temp-to-hire conversion fees of 10% to 25% of first-year salary with waiver thresholds commonly near 520 to 1,000 or more worked hours (vendor surveys); direct placement fees of 15% to 25% with 20% the common benchmark (Top Echelon 2025). Survey data, labeled as ranges, never as your quote. altline.sobanco.com, staffing markupChecked 23 July 2026
The statutory lines are current federal figures as of the dates shown; the markup, buyout, and fee figures are published survey ranges, not quotes, and your agreement controls. State unemployment and workers compensation rates are specific to your state, experience rating, and class codes, and they change. General information, not legal or tax advice; confirm your own rates and terms before building a business case on them.
Tools that run this math
Model the cost first, then close the risk
Common questions
Published 2025 survey ranges run 25% to 75% of the worker pay rate, varying by segment, volume, and risk class. Light industrial and clerical placements at volume sit near the bottom of the band; specialized and low-volume placements run higher. Your agreement is the number that matters, and a markup transparency clause is how you see it.
No. The markup first funds the statutory burden the agency carries as the employer of record, then screening, onboarding, replacement, and the agency overhead, and only then margin. The savings case for conversion is the burden and margin you stop renting, not a claim that the markup is pure profit.
Build it from the lines: 7.65% employer FICA, an effective 0.6% FUTA, your state unemployment rate from your rate notice, your workers compensation class rate, and a benefits load near 30.1% of total compensation if you extend benefits. An 18% statutory-only figure and a 30% with-benefits figure are reasonable planning defaults, and both should give way to your real rates.
When the remaining need is long, the buyout is waived or low, and your loaded burden sits well under the agency markup. Short remaining need, a high buyout, or a loaded burden near the markup flips the answer to the agency. And a conversion into independent contractor status instead of employment is a classification decision, not a cost decision; run the tests and involve counsel before taking that route.