The two hidden assumptions
Salary divided by 2,080 gives an employee's hourly cost as the employer sees the payroll line, and only that line. It is useful for exactly one purpose: comparing two salaried offers. As a freelance rate it fails twice.
The first failure is billable hours. A salaried employee is paid for time in a planning meeting, writing an internal document, learning a tool, waiting on someone else. A freelancer does all of that and bills for none of it. Add the work unique to being independent — finding clients, writing proposals that go nowhere, scoping, invoicing, chasing invoices, bookkeeping — and the billable share of a working week lands, for most people, between half and two thirds. Forty desk hours is realistically 20 to 26 billed.
The second failure is cost. The employer was paying for things that never appeared on the paycheck: their share of payroll taxes, most of a health premium, a retirement contribution, equipment, software, training and paid time off. Every one becomes a line in your own budget the day you go independent.
A rate, built from the bottom
Take someone leaving an $85,000 salaried role who wants to be no worse off. Every figure below is illustrative — yours will differ, and some come only from your own quotes and your own accountant — but the shape transfers.
Step one: how many hours can you actually bill? Assume 40 desk hours a week. Take three weeks of vacation, a week and a half of holidays, and two weeks of illness and dead time between contracts: about 45 working weeks. Assume 55 percent of desk time is billable — realistic for an established solo practice, optimistic for someone starting out. That is 22 billable hours a week.
45 weeks × 22 hours = 990 billable hours a year. Not 2,080. That number alone doubles the naive rate before a single cost is added.
Step two: what did the employer stop paying for?
| Cost the employer used to absorb | Illustrative annual | Notes |
|---|---|---|
| Health coverage for the household | $9,600 | Varies enormously by state, age, household size and plan. Get a real quote for your own situation — this is the line most likely to be wrong by thousands in either direction. |
| Retirement contribution you now fund entirely | $6,000 | Replacing an employer match plus your own savings. Which account types are available to self-employed people, and what the limits are, is a CPA question. |
| Additional payroll-tax burden | $6,500 | An employee splits payroll taxes with their employer; a self-employed person carries both halves. The amount depends on income, structure and state, and there are offsetting deductions. Estimate it with a CPA — the rates and thresholds change and are not worth guessing at. |
| Equipment, software, subscriptions, phone and internet share | $2,400 | Machine replacement amortized, tools, hosting, storage |
| Business insurance and professional fees | $1,800 | Liability or errors-and-omissions coverage where your field expects it, plus accounting and any filing costs |
| Subtotal | $26,300 | None of this appeared on the old paycheck, and all of it is now yours |
Step three: put it together. Target income $85,000 plus $26,300 of costs is $111,300. Add a buffer of about 5 percent for bad debt, underestimated projects, and the client who disappears mid-invoice: call it $117,000 of revenue needed.
$117,000 ÷ 990 billable hours = about $118 an hour.
Against the naive $40.87, that is nearly three times. Neither is a market rate — the market decides what your work is worth and may not support $118 — but $118 is the honest cost of doing that work at that income, and knowing it changes what you accept. A project at $55 an hour is not a slightly worse deal than $118; it is a pay cut of more than half against the salary just left, dressed as self-employment. The freelance rate calculator runs this on your own figures, and the hourly value calculator does the reverse check: what a job actually paid once you count the hours it really took.
A floor and a target are different numbers
Two figures belong on paper before you quote anything. The floor is the rate below which you lose money: total costs divided by billable hours, ignoring income entirely — $26,300 ÷ 990, about $27 an hour here, and anything below it is subsidized by you. The target is the $118. Between them is a band where you might rationally accept work, but you should know when you are in it and why, rather than discovering it in December.
Then know your utilization honestly. Track hours for two months, including everything unbillable. Most people find their billable share is lower than the 55 percent assumed here, and at 45 percent the required rate rises by about a fifth. The work hours calculator and work time log tally do the totaling.
Hourly, fixed, retainer, value
| Model | Who carries the risk | Works when | Fails when |
|---|---|---|---|
| Hourly | The client | Scope is genuinely unknown, ongoing support, work that changes direction | Clients dislike an open meter, it caps your income at your available hours, and it penalizes you for getting faster — the better you get, the less you earn for the same result |
| Fixed price | You | Scope is clearly bounded and you have done similar work enough times to estimate it | Scope moves. This is the central risk of freelancing and it is why the scope and revision clauses matter more than the price. |
| Day rate | Shared | Blocks of dedicated work, on-site engagements, anything where the client wants your whole attention | Half-days and fragmented weeks. Define what a day is before you quote one. |
| Monthly retainer | Shared | Ongoing needs, and it smooths the income variance that makes freelancing stressful | Scope drifts upward until the retainer buys twice what it did. Needs a defined ceiling and an annual review. |
| Value or outcome pricing | Mixed | You can point at a number the work moves and the client agrees how it is measured | Most of the time. Requires trust, a measurable outcome, and a client who will not dispute attribution later. |
Clients prefer fixed price because it gives them a known number, and what you are selling them is exactly that: certainty. You absorb the risk that the job runs long, and you should be paid for carrying it. Quote at your honest hour estimate times your rate, then add 15 to 30 percent of contingency depending on how well you know the work and the client. A fixed price with no contingency is a bet you lose more often than you win, because estimates are asymmetric — jobs run over far more often than under.
Scope creep, with the arithmetic attached
A design job is quoted at $6,000 for an estimated 60 hours: an effective $100 an hour, comfortably above the floor. The client asks for a variation on the homepage, then a different color direction, then a version for a stakeholder who joined late, then wants the first version back. None of these is unreasonable on its own and none feels big enough to argue about. They add 25 hours.
$6,000 ÷ 85 hours = $70.59 an hour. A 30 percent pay cut delivered in four polite emails, and the freelancer usually notices only after the project closes.
The defense is not firmness in the moment, which is hard and makes you the difficult party. It is a number written down before anything starts: the deliverable, how many revision rounds are included, what counts as a round, and the hourly rate for anything beyond — agreed in advance, so the conversation is administrative rather than adversarial. "Happy to do that. It falls outside the two included rounds, so about six hours at the change rate we agreed — go ahead?" is a sentence nobody resents, and it is only available if the number exists. What a contract should settle is in the freelance contract guide; a quotation form is a reasonable place to make the scope legible from the start.
Raising rates on existing clients
Freelance income stagnates for a structural reason: new clients get current rates while old ones keep paying what they agreed two years ago, so the longest relationships become the worst-paid. Left alone, this resolves only when you resent your best client.
What works is being unremarkable about it. Give 60 to 90 days of notice, state the new rate and its start date, do not apologize, over-explain, or offer a menu of alternatives. Raise everyone on the same schedule so it is policy rather than a judgment about them. Expect a few departures and price that in: if a 12 percent increase costs you the least profitable fifth of your clients, you are usually ahead on both money and hours. The price increase simulator shows how much volume a given increase can afford to lose, and the answer is normally more than instinct suggests.
The things that are not a rate problem
Two failure modes look like underpricing and are not. The first is income variance: a good annual total delivered lumpily still bankrupts someone with no reserve, which is why independent workers need a larger cushion than employees — see emergency fund sizing and the emergency fund calculator. The second is tax. Nobody withholds on your behalf, what you owe and when depends on your structure, state and situation, and the standard disaster is spending money in June that belonged to a tax bill. Set it aside in a separate account as it arrives, at whatever share your CPA tells you, and treat that account as not yours.
One more: whether an engagement is properly a contractor relationship or an employment one is a legal question with real consequences for both sides, and the tests differ between federal and state law and between agencies. If a client sets your hours, supplies the equipment, directs how you work and treats you as staff, ask an employment attorney or a CPA rather than assuming.
Set the rate from your own arithmetic, then let the market argue. A rate you cannot explain is one you will discount the moment someone pushes.
Questions people ask
How do I convert a salary into a freelance hourly rate?
Not by dividing by 2,080, which is roughly a third of the real answer. Work it from the bottom instead. First, count billable hours: subtract vacation, holidays, illness and gaps between contracts to get about 45 working weeks, then assume only half to two thirds of desk time is billable, because proposals, scoping, invoicing, admin and learning are unpaid. Forty desk hours a week is commonly 20 to 26 billable, so call it around 990 billable hours a year. Second, add the costs your employer used to absorb — health coverage, retirement funding, the additional payroll tax burden a self-employed person carries, equipment and software, insurance and accounting. Third, divide. In a worked example an $85,000 salary becomes roughly $117,000 of needed revenue over 990 hours, or about $118 an hour against a naive $40.87.
How many hours a week can a freelancer actually bill?
For most solo practitioners, somewhere between half and two thirds of the hours they spend working — so 20 to 26 billable hours out of a 40-hour week, and lower than that in the first year or two while you are still building a client base. The unbilled remainder is real work: finding clients, writing proposals that do not convert, scoping, invoicing and chasing payment, bookkeeping, learning tools, and the ordinary administrative overhead an employer used to handle. The practical step is to track everything for two months, including the non-billable time, and compute your own percentage rather than using anyone else's. If your billable share turns out to be 45 percent rather than 55, the rate you need rises by about a fifth, and that is not a rounding error.
Should I charge hourly or a fixed price?
The real question is who carries the risk that the work takes longer than expected. Hourly puts it on the client, which they often dislike because the total is unknown, and it caps your income at your available hours while penalizing you for getting faster. Fixed price puts the risk squarely on you, which is why clients like it: what they are buying is certainty, and you should be paid for supplying it. Quote a fixed price at your honest hour estimate times your rate, then add 15 to 30 percent of contingency depending on how well you know the work and the client, because estimates are asymmetric — projects run over far more often than they run under. Whichever you choose, the clause that actually protects a fixed price is not the price, it is a written scope with a defined number of revision rounds and an agreed rate for anything beyond them.
How do I handle scope creep without being difficult?
By making the answer administrative instead of personal, which requires a number that exists before the work starts. Write down what the deliverable is, how many rounds of revision are included, what counts as a round, and what additional work costs per hour. Then the conversation becomes "happy to do that — it falls outside the two included rounds, so it would be about six hours at the change rate we agreed, shall I go ahead?", which nobody resents because it was decided in advance and applies to everyone. Without that, each request is individually reasonable and collectively expensive: a $6,000 fixed project estimated at 60 hours is $100 an hour, and 25 hours of polite unbudgeted revisions turns it into $70.59, a 30 percent pay cut delivered in four emails.
How do I raise rates on a long-standing client?
Treat it as policy rather than as a negotiation about them specifically. Give 60 to 90 days of notice, state the new rate and the date it takes effect in two or three sentences, and stop there — do not apologize, do not over-explain, and do not offer a menu of alternatives, all of which invite a negotiation you did not want. Raise everyone on the same schedule so it is clearly a rate change and not a judgment. Expect to lose a few clients and decide in advance that this is acceptable: run the arithmetic first, because a modest increase can usually afford to lose more volume than instinct suggests, and the clients who leave over a 12 percent rise are typically the least profitable ones anyway. Freelance income stagnates mainly because new clients get current rates while old ones keep paying rates set years ago.