FTE vs pay per return outsourcing reads like a price question, and it settles something else: who pays for the weeks when the work is thin. A dedicated seat bills every month whether or not you fill it. A per return price bills only when a return moves.
Both can be the cheaper choice for the same firm in different years. What separates them is your own break-even return count, and you can work that out from the two quotes already in front of you.
Work through the break-even, the season test, and the contract terms, and you will have your answer before either provider follows up.
What FTE and Pay Per Return Mean in an Outsourcing Contract
An FTE, short for full-time equivalent, is a contracted seat. One person, or an agreed fraction of one, works your files for a set number of hours each month at a fixed monthly fee. You are buying availability, and filling the seat is your job.
Pay per return prices output instead. You are quoted a fee per return by form type and complexity tier, and you pay when work goes out. Send nothing in June, pay nothing in June.
Almost everything else in the comparison follows from that one difference. It sets the break-even, the terms worth negotiating, and which side of the table absorbs a quiet February.
One naming clash is worth clearing up early. Pay per return is also a licensing option in professional tax software, where you authorize payment for each return as you process it rather than buying a full seat license. Same phrase, different purchase: one buys software rights, the other buys labor.
Who Carries Which Risk
The two models split four risks between you and the provider, and that split is the real product difference.
Idle capacity. Under an FTE you carry it. The fee runs through the quiet months, and whether unused hours carry into the next one is a term you negotiate, not a default you inherit. Under per return the provider carries it and prices it into the unit rate, so a seasonal firm is choosing between paying for its own trough and paying a premium that covers someone else's.
Peak volume. A seat has a ceiling of hours. Past that you queue, buy overtime, or add a second seat in the worst possible weeks. A per return provider draws on a pool, which is capacity you do not own, so your work gets sequenced against other firms' deadlines.
Complexity drift. An FTE absorbs a harder return inside hours you have already bought. Per return re-prices it, so a mix drifting toward multistate and entity work raises your unit cost without anyone reopening the contract.
Knowledge of your files. A named seat learns your workpapers, your clients, and what your reviewer sends back, then takes all of it with them if they roll off. A pool spreads your work across preparers who each start further back, which shows up as questions and rework rather than as a line on the invoice.
Set against each other, the split reads like this.
| Risk | Under an FTE | Under pay per return |
|---|---|---|
| Idle capacity | Yours, the fee runs anyway | The provider's, priced into the rate |
| Peak volume | Capped by the seat's hours | Pooled, sequenced against other clients |
| Complexity drift | Absorbed in hours you bought | Re-priced by tier |
| Knowledge of your files | Concentrated in one person | Spread across a pool |
FTE vs Pay Per Return Outsourcing: Find Your Break-Even Return Count
One division gets you most of the way there. Divide the monthly seat fee by the per return price for the same work, and you have the number of returns a month at which the two quotes cost the same. Send fewer than that, per return wins the month. Send more and the seat wins, up to the point where the volume no longer fits inside the seat's hours, after which you are comparing a second seat against the per return rate.
``` Break-even
seat fee per month
-------------------------- = returns per month where the two quotes tie price for that return
Then compare month by month, across the term you would sign:
FTE month seat fee + your review hours + your admin hours Per return month (returns sent x price per return) + your review hours + your admin hours
Count the off-season months too. A seat bills in July. Per return does not. ```
Two conditions make that division honest. Both quotes have to describe the same deliverable, because a return that comes back reviewed and tied out is a different product from one that comes back keyed. And both columns need your own retained hours in them, since neither model takes your reviewer out of the chain.
Price those retained hours, and say which rate you are pricing them at, because what the reviewer's hour costs you and what you could have billed it for produce different break-evens. Pick one and use it in both columns. The 2025 National MAP Survey put the median net hourly billing rate at $170, up from $159 two years earlier. That number tells you whether a billing-rate assumption sits in the right range. It says nothing about what the hour costs you.
For the cost side there is a federal figure to check yours against. The median hourly wage for accountants and auditors was $40.23 in the May 2025 Occupational Employment and Wage Statistics survey. That is wage alone, before anything an employer adds on top of it, so a reviewer hour costs you more than the wage line whichever number you start from.
Run it on your real mix, not on one form type. A per return price moves with the form while the seat fee does not, so price every return you would send at its own tier, add those up, and set that total against the single seat fee. A mix drifting toward entity work crosses the seat fee at a lower return count than the individual-return math suggests.
What Your Season Shape Does to the Answer
Look at when your returns move before you commit to a model. Tax professionals had e-filed 12,724,000 returns as of February 13, 2026, and 72,821,000 as of April 17, 2026, according to IRS filing season statistics.
The tail after the deadline is thin. Three weeks on, the professional e-file count stood at 75,316,000 as of May 8, 2026.
A seat you pay for in twelve monthly installments is fed by work that arrives in a burst. That is not an argument against the seat. It is an argument for knowing what the seat will do between May and December before you sign for a year.
Firm survey data points the same way. In the 2025 National MAP Survey, 59% of respondents put the share of their work handled by an outsourced or global team at 1% to 5%, and another 25% put it at 6% to 15%. Of the firms doing it, 72% used the vendor model, which relies on a third party, rather than an employer-of-record model or a wholly owned offshore facility. A slice that size, bought from a third party, is the volume your break-even has to clear, and it does not always fill a seat.
So write the off-season list before you run the season arithmetic. Monthly bookkeeping, payroll, extensions, notices, cleanup work, and workpaper standardization are what keep a seat busy in the quiet months. If that list will not fill the hours you are being asked to buy, you are pricing a year and using a quarter.
Where Each Model Hides Its Rework
Rework is paid for in both models. What changes is whether you can see it.
Under an FTE, rework never appears on an invoice. A preparer who redoes a schedule is spending hours you already paid for, so the cost lands as a seat that produces less than you planned. The only way to see it is to count returns delivered against hours billed, month by month, from the first month.
Under per return, rework has a price, and the contract decides who pays it. Ask whether a return sent back for correction is fixed under the original fee, billed again, or capped at a number of passes. A provider who will not answer that in writing has left an open line item in your budget.
Either way, the number that settles the argument is what a finished, accepted return costs you once your own review time is counted. Two providers at the same headline rate can sit a long way apart on that.
The Terms That Decide Whether the Quote Holds
A quote is only as good as its definitions. These are the terms worth pinning down before either model is priced.
If You Are Buying an FTE
- Hours in an FTE month. Get the number in writing, along with the working calendar and the public holidays that apply where the team sits. Two seats quoted at the same fee can differ by several days of output a year.
- Overtime and surge. Ask what happens when your peak needs more than the seat's hours, at what rate, and with how much notice. That clause is what your March looks like.
- Unused hours. Find out whether hours roll forward, expire at month end, or can be banked into the season. It changes the real cost of your quiet months.
- Ramp and replacement. Establish how long the seat takes to become productive, who pays for that time, and what happens if the person is not a fit or rolls off mid season.
- Named people. Ask who exactly will work your files, and how you are told when that changes. A seat that is quietly reassigned is a pool with a seat's price tag.
If You Are Buying Per Return
- The definition of a return. Settle whether state filings, extensions, amendments, and consolidated entities are separate charges. This single definition moves a per return budget more than the headline rate does.
- Complexity tiers. Get the tier boundaries written as tests you can apply yourself, not as adjectives. Otherwise your returns drift upward through the tiers as the season gets busier.
- Rework policy. Fix the number of correction passes included, and what a return has to look like when it comes back for the fee to be earned.
- Turnaround in the peak. Ask for the committed window in the busiest weeks specifically, since off-season turnaround tells you nothing about April.
- Minimums and price validity. Check for monthly or seasonal minimums, which quietly convert per return into a seat, and confirm how long the quoted prices hold.
What the Pricing Model Does Not Change
Your duties as the preparer of record are the same under both quotes. Two of them decide how an offshore engagement has to be set up, whichever column you sign.
One preparer may pass tax return information to another preparer without consent only in narrow circumstances. The regulations authorize disclosure to another tax return preparer located in the United States for preparation or auxiliary services, so long as the services provided are not substantive determinations or advice affecting the tax liability reported by taxpayers. Once the preparer receiving that information sits outside the United States, the taxpayer's written consent is required before the disclosure.
The Social Security number is the piece to settle before either quote is signed. On a return in the Form 1040 series, the preparer inside the United States must redact or otherwise mask the number before the information goes offshore. The one exception the regulation allows runs the disclosure through an adequate data protection safeguard, with the consent itself verifying that the safeguard is kept in place. Neither a seat nor a unit price buys you out of that step, so ask each provider which of the two routes they operate.
Your security program travels with the work as well. The FTC Safeguards Rule treats an accountant or other tax preparation service that completes income tax returns as a financial institution, and selecting and overseeing service providers is one of the required elements of the written information security program it demands. That element asks you to take reasonable steps to select providers capable of maintaining appropriate safeguards, to require those safeguards by contract, and to reassess the provider periodically.
The duties do not shift with the pricing model, but your ability to evidence them does. With a named seat you can answer "who worked this file" from a short list. With a pool you cannot, unless the contract obliges the provider to tell you, so settle that before the first file moves.
When Each Model Is the Wrong Choice
An FTE is the wrong first move when you cannot name the off-season work, when the provider is unproven with your firm, when your volume swings in ways you cannot forecast, or when your processes have never been written down. Buying a seat before those are settled means paying by the month for someone's learning curve.
Pay per return is the wrong long-term answer when the work needs to know your clients, when you need someone reachable the same day, when files are messy enough that the unit price gets renegotiated every few weeks, or when your volume has been steady for two seasons and you are still paying a premium each month for flexibility you no longer use.
There is also a case for neither. If what you need off your plate is judgment work, the positions, the elections, and the client conversations, neither model helps. That is a hiring or partnership decision in disguise.
Questions Firms Ask
What Is an FTE in Outsourcing?
An FTE is a unit of contracted capacity, not a job title and not an employee of your firm. What it means in practice is whatever the agreement says about hours, calendar, and coverage, which is why the hours definition matters more than the label.
Is a 0.75 FTE Considered Full Time?
By the federal hours test, yes. Under the employer shared responsibility rules, the IRS treats a full-time employee as one employed on average at least 30 hours of service per week, or 130 hours of service per month, so three quarters of a 40-hour week is 30 hours and lands exactly on that line. A second federal rule counts differently. For the small business health care tax credit, the weeks-worked equivalency credits 40 hours of service for each week paid, and FTEs are counted by dividing total hours of service by 2,080.
Both of those describe people on your payroll, and neither governs what a vendor puts in a proposal. A three-quarter seat bought from a provider is a fraction of a contract, so it means the hours the contract names and nothing more.
Is Pay Per Return Cheaper Than an FTE?
Below your break-even count, yes. Above it, no. Work the count out per form type, then check it against the volume you actually sent each month last season rather than the volume you hope to win.
Can a Firm Use Both?
Yes, and the split should follow which work is predictable rather than which is cheap. Hold a seat for the volume you can name twelve months out, send everything above it per return, and write the dividing rule as a test rather than a description, naming which return types, above what monthly count, at which tier. Without that test, work you thought the seat already covered comes back as a per return line.
Start With the Count, Not the Rate
Take the two quotes, divide the seat fee by the per return price, and compare that break-even against the months you can actually document from last season. Then add your review hours to both columns and count the off-season months at their real value. The model that survives that comparison is the one to sign.
Do the arithmetic in a quiet month, when a wrong estimate costs you nothing but the time it takes to redo it. Then prove the winner on a small block of live work before a season depends on it.
Accountably places trained offshore accountants and tax preparers inside US CPA and EA firms, ramped on your software and SOPs in about 3 to 4 weeks. Since 2022 that has come to 30+ placements across 20+ US firms. If you would rather grade the work than model it, the Free 40-Hour Proof Pilot puts a fixed 40-hour block of your own representative work through multi-layer review, so your reviewer sees real output before your name is on the line. Don't trust us. Test us.