Paraform publishes a payout example and never publishes a denominator

Paraform's recruiter page carries an interactive earnings example. In the state it loaded in on 4 September 2026 it showed a success fee split of 80%, alongside $15k per month and five hours a week of work on the platform.

Two more figures come from Paraform's own blog. The Series B announcement of 18 March 2026 gives average candidate total compensation of $260,000 a year. A comparison post states the company-side fee: "roughly 25% of first-year salary, and you pay nothing until someone accepts an offer." Those two figures do not multiply.

The fee runs on first-year salary, and the company-facing page quoted below narrows that to base salary only. The $260,000 is total compensation, platform-wide, in a post that also puts packages at $300K to $400K. Paraform publishes no average base salary anywhere. So what its own numbers support is this: a won placement pays the recruiter 80% of 25% of a base salary Paraform does not disclose.

Put an illustrative $200,000 base through that, ours and not Paraform's, and the fee is $50,000 and the recruiter's share $40,000. Every dollar estimate of Paraform earnings, this one included, is running a fee percentage over a salary figure nobody has published.

No Paraform page says how many searches you worked to get there. The platform publishes no win rate, and no count of how many recruiters can work one requisition at once. That was checked on 4 September 2026 across the recruiter and company pages, the recruiting agreement, the ground rules, the charter, the help-centre FAQ, four blog posts and /pricing, which returns a 404.

The number on the marketing page, and the number missing from it

Be precise about what that example is. It is a slider-style widget with a default state, not a sentence in the page's prose, and Paraform does not call the 80% its policy or the $15k figure an average, a median or an audited result.

Compute forward anyway, because two of the three inputs are Paraform's own. The 25% is not published where anyone would look for it. It appears once, inside a post comparing Paraform to a competitor, while the company-facing page gives the model and no number: "a success-based model - we receive a flat percentage of first year base salary only upon hire," plus a 90-day free replacement search. The pricing URL is a 404.

That quoted clause is also what rules out the obvious shortcut. First year base salary is the fee base. The $260,000 is total compensation. Paraform publishes the second and not the first, so the honest statement of a win is 25% of an unpublished base, of which 80% reaches the recruiter.

Every dollar amount past that point carries a base salary chosen by whoever wrote it. On an illustrative $200,000 base that is a $50,000 fee and $40,000 to the recruiter. On $150,000 it is $37,500 and $30,000. Keep that mismatch in view through the rest of this piece: it is the flaw in every third-party dollar figure for Paraform earnings, and it would have been the flaw in ours.

The $260,000 total-compensation figure carries a second caveat Paraform supplies itself. It is a platform-wide average across every role and recruiter. The same Series B post describes an uneven offer distribution: the top 12 percent of candidates receive more than a quarter of all offers, while the bottom 40 percent receive roughly the same share between them. An average sitting on a distribution that skewed does not describe the requisition you are about to pick up.

One company-side charge is confirmed historically and not confirmable as current. In April 2024, TechCrunch quoted Paraform's CEO describing the model as "a listing fee (subscription fee) to publish jobs on the platform and a success fee when a hire is made," with no percentage or dollar figure attached. No Paraform page read on 4 September 2026 repeats it, the company-facing page describes only a success fee, and /pricing is a 404, so whether a listing fee is still charged in 2026 could not be confirmed either way from public pages. Dover, which sells an alternative to Paraform, asserts it as current pricing and cites nothing.

Now the figure a recruiter searching for reviews meets before any of this, and the reason it appears below only as a labelled contrast. Google's AI Overview for paraform review, read on 4 September 2026, attributes to Glassdoor a payout of "$10,000-$15,000 average per successful placement." That range could not be located on any Glassdoor page in this search. It appears close to verbatim in a review post by Nextdev, a competing hiring platform, which cites no source of its own, and it sits 2.7 to 4 times below the $40,000 that Paraform's own fee and split imply on the illustrative base above.

Sentiment does not fill the gap either. Read on 4 September 2026, Trustpilot showed 2.5 from 9 reviews and Glassdoor 3.7 from 26 reviews, samples too small to support a conclusion about a company. FeaturedCustomers showed 4.8 from 732 reviews whose reviewers are named client logos: companies that hired through Paraform, not recruiters who worked its requisitions.

What Paraform actually pays, and when

The payout structure sits in the recruiting agreement, the document that binds you and is drafted by the counterparty. Payment arrives in three installments, one third at 30 days, one third at 60 days and one third at 90 days, and the clock starts at the candidate's start date rather than the signed offer.

The employer, under the same agreement, pays Paraform within 14 business days of the candidate signing. So the platform is paid in full before the recruiter is paid at all. What it holds after that declines. It keeps the whole of the recruiter's share through the notice period and the first 30 days after the start date, two thirds from day 30, one third from day 60, and none from day 90.

The gap is not only timing. Inside a 90-day guarantee period, the agreement provides that "all or some of the referral bonuses...will be refunded if Candidate leaves, or is terminated." A separate clause gives the recruiter 60 days to place a replacement. Fail that, and 100% of placement fees already paid out must be repaid. Repayment is enforceable by automatic deduction from future payouts or by wire within 7 calendar days, escalating to third-party collections on non-compliance.

The exposure is a range rather than a ceiling, because only fees already paid out are repayable and payout arrives in thirds. On the illustrative $40,000 win, a candidate who leaves before day 30 leaves nothing to claw back, though the remaining installments stop. From day 30 the amount at risk is $13,333, from day 60 it is $26,667, and the full $40,000 is exposed only at the boundary, where the third installment lands on day 90 and the guarantee window closes on day 90. The 7-day wire is one of two enforcement routes named, the other being deduction from future payouts.

Two ownership rules decide whether sourcing hours turn into anything. Duplicate submissions are settled by timestamp, so a candidate you found first and submitted second belongs to the other recruiter. And a candidate already in an employer's pipeline before you submit belongs to the employer, so a pipeline you cannot see can void your outreach.

On exclusivity, what the agreement says is narrow rather than absolute. The recruiter is styled an independent contractor, "free at all times to provide services to persons or businesses other than Paraform," so there is no platform-wide exclusivity clause.

Read the narrower restrictions twice. No direct engagements with employers you found through Paraform while you are active, no soliciting those employers' staff for 6 months after your engagement ends, and a 12-month non-solicitation of the employers themselves after account termination. The ground rules attach the consequence, "immediate suspension or termination" plus forfeiture of any unpaid payments. They add conduct rules with teeth: do not present yourself as "a [Company Name] recruiter," do not inflate compensation ranges, and do not submit a candidate without their consent, which carries immediate termination.

So this is contingency work in which the client relationship is not yours to keep, and the non-solicitation clause is the price of not having to find a client. Recruitment agency fees lays out what contingency, retained, engaged and contract work charge, and when each gets paid.

The bar that keeps your account in good standing

The Charter and Guidelines sets a performance floor, evaluated on a rolling ten submissions. An interview rate above 25%. An average rating of at least 2.5 out of 4 across your submissions, with no more than one submission rated 1 out of 4. And no gap longer than 30 days between submissions.

Each of those carries a cost the payout example does not. A rolling ten-submission window means a quiet month still counts against you. The 30-day rule prices patience: you cannot work two requisitions carefully and sit out the rest of the quarter, which is the behaviour a recruiter uses to protect an hourly rate. And an interview-rate floor pushes submissions toward requisitions whose answer you already know.

Then there is a discrepancy in Paraform's own documentation, and it is worth more than any review. A separate FAQ page states a recommended target of a 60% first-round interview rate and a 30% mid-round interview rate. The Charter's account-standing floor is 25%.

Those are different numbers on the same axis, in two documents on the same company's site, and nothing found on either page reconciles them. A recruiter deciding whether their account is safe cannot tell from the published material whether 25% or 60% is the number that matters. That is a documentation gap, not an allegation.

Effective earnings per hour, three scenarios

What follows is a scenario model, not a measurement. It blends figures Paraform publishes with an hours assumption imported from a different context and a concurrency range nobody has measured. Every input is named, with its source, before any result appears.

One of those sources is us, so our own interest gets the same label this piece puts on Dover and Nextdev. Glozo sells recruiting software into this market, the hours and the hourly floor below are our own figures, and the piece ends in our own CTA.

Ninety hours on a search that fills and 40 on one that dies come from the solo desk P&L, which models a generic US solo recruiting desk. They are hypothetical there and stated as such: that piece asks the reader to say a filled search takes 90 hours and works forward from the supposition. They are planning figures rather than measurements, they are not Paraform figures, Paraform publishes none, and a marketplace requisition may consume a different number of hours. They are here because a model with no hours in it produces no hourly rate.

With those, the hours to reach one win are 90 + 40 × (N - 1), where N is how many recruiters are working the requisition.

No row in that table is the typical one. It would be dishonest to nominate one. N is the variable the answer turns on, Paraform does not publish it, and no third party measures it credibly. The six cells show a spread, and the spread is the finding.

Read them against a floor. The solo desk model prices a US recruiter's own loaded hour at $45, from the $91,000 loaded annual figure that piece builds out of Glozo's own market data. That model is precise about what the number is: the floor beneath which that desk is not worth running, to be used as the cost of your labour rather than as its value.

At the illustrative $40,000 win, every scenario clears that floor comfortably, even at ten recruiters per requisition. At the untraceable $12,500 figure, the same model clears it at three recruiters, roughly breaks even at five, and falls below a salaried recruiter's loaded wage at ten.

Both columns are uncertain, for different reasons. The high one rests on numbers Paraform publishes in places that read as illustration rather than policy, and on a base salary we chose because Paraform publishes none. The low one rests on a figure whose attributed source does not carry it.

What survives either way: the marketing example prices a win and never a loss, and no Paraform page states a number from which a reader could price the losses. Contingency versus retained recruiting runs the same shape of calculation, fee times probability over the hours the work consumes whether you win or not.

Where a marketplace beats a desk, and where it does not

Start with what the arithmetic concedes. The fee base looks high, though Paraform does not publish it. A platform-wide average candidate total compensation of $260,000, "often reaching $300K to $400K" in Paraform's own words, is an equity-inclusive figure rather than a salary. The 25% fee sits inside the 15 to 30% range the market pays for contingency work, as recruiter salaries for freelance and agency desks shows.

The Series B post reports over $50 million paid to recruiters to date, 1,000+ companies, and an average time to a candidate meeting of about 12 days. Business development is real money and real hours on a solo P&L. A marketplace deletes that line.

Now the side the payout example leaves out. You do not own the client, and the non-solicitation clause means you cannot convert one, during your engagement or for 12 months after termination. Your cash arrives in thirds starting 30 days after someone's first day, with a refund window over it and a 60-day replacement obligation behind that. Your right to keep submitting rests on a rolling ten-submission bar you did not set, and Paraform publishes no count, anywhere on its own pages, of how many recruiters are working a requisition.

That is a specific trade, not a better deal or a worse one. A marketplace requisition converts client-acquisition risk into competition risk, and competition risk is the one you cannot measure from outside. A recruiter with three repeat clients knows their win rate because they have one. A recruiter working an open marketplace requisition against a number of others Paraform does not publish is pricing a search without the denominator.

It is also why a nine-review page cannot settle this: the same payout reads as generous at three recruiters per requisition and thin at ten. For the same decision from an employer's side, in-house versus freelance recruiting covers who carries which cost.

What to ask before you list your first search

Price your own hour first, because every question below is only answerable against a number. The solo desk model puts a US recruiter's loaded hour at $45: the floor beneath which the desk itself is not worth running, because you could earn it working for somebody else. Use it as the cost of your labour, not as its value.

Then ask Paraform, in writing, the four things its pages do not answer. How many recruiters are working this requisition right now, and does that change once I start. What is the split on this requisition, and where is it stated other than in the earnings example. Which interview rate governs my account, the Charter's 25% floor or the FAQ's recommended 60%, given that the two do not reconcile. And if a placement falls off inside the guarantee window and I cannot deliver a replacement in 60 days, what is deducted, from which future payouts, and by when.

Two more before the first submission. Confirm when your payout clock starts, because the agreement starts it at the candidate's start date rather than the signed offer, so a long notice period is your working capital. And confirm what a candidate already sitting in the employer's pipeline does to your submission, because the agreement gives that candidate to the employer and those hours are yours to absorb.

Then keep the record the platform does not publish. Log hours per requisition, submissions, interviews and wins from your first day. After ten submissions you hold your own win rate, which is the one number that turns the table above from a scenario into a measurement of your desk. It will not tell you how many recruiters you were up against on any given requisition, because nobody publishes that, but it prices the outcome of the competition without needing to count it.