How do you price freelance work?

You price freelance work by building a minimum rate from the bottom up, then choosing how to present it to the client. Add the income you need to keep, the tax and social charges on it and the cost of running the business, divide that total by the hours you can really bill in a year, and you have the rate below which the work loses money. Hourly, daily, per-project and value-based pricing are four ways to sell above that floor, and each one moves the risk of an overrun to a different side of the table.

The division is where underpricing usually starts. A freelancer who divides a salary by a full-time year of hours assumes every working hour is paid, and none of them are when you are writing a proposal or chasing an invoice. The method below computes the rate once, with one worked example carried from the first line to the last, and then uses it to estimate a project, to decide when a range is the right answer, and to measure what a discount or an unbilled revision does to the rate you actually earn.

Every amount in the example is illustrative and in US dollars. Tax and social charges differ from one country to the next, so the example uses a single stated percentage that you replace with your own. This is general information, not tax advice.

What does a freelance rate have to cover?

A freelance rate has to cover four things an employee never sees on a payslip: the income you keep, the tax and social charges on your profit, the costs of running the business, and the working time no client pays for. Leave one out and the rate looks competitive while the year closes short.

The first item is the only one most people start from. The other three are what an employer handles out of sight. An employee's pay is only part of what the job costs: according to the US Bureau of Labor Statistics, benefits made up 30.0% of employer compensation costs for private industry workers in June 2026, with wages and salaries the other 70.0%. Paid leave, insurance, retirement contributions and the employer's share of payroll taxes sit in that 30.0%. When you work for yourself, each of those lines becomes yours to fund, and the rate is the only place the money can come from.

Tax and social charges change shape as well as size. In the United States, for instance, the Internal Revenue Service sets the self-employment tax rate at 15.3%, made of 12.4% for social security and 2.9% for Medicare: a self-employed person pays both the employee's and the employer's share, before income tax is counted. Other countries collect social contributions through different mechanisms and at different levels, and some apply a flat percentage to turnover under simplified regimes. The method does not depend on the mechanism. It needs one number from you or your accountant: the share of your profit that leaves in tax and social charges, taken as an effective rate over the whole year.

How much revenue do you need to bill in a year?

The revenue you need to bill in a year is your target take-home income, divided by one minus your effective tax rate, plus your business costs. Work backwards from what you want to keep, not forwards from what a client might pay.

Take a freelancer who wants $72,000 a year left after tax and after costs, which is $6,000 a month. That target should already include what an employer used to fund quietly: retirement saving and a cushion for weeks without work. Assume tax and social charges take an illustrative 28% of profit. Keeping $72,000 then needs a profit of $72,000 divided by 0.72, which is $100,000, and $28,000 of it goes to tax and social charges.

Business costs come on top, because they are paid before there is any profit to tax. List them from last year's bank statements instead of from memory: subscriptions and occasional purchases are the lines people forget.

Spread evenly, $116,000 is about $9,667 of invoices a month, every month, including the ones with a holiday in them. A quarter that bills less than three times that amount has to be made up later.

Two cautions keep this side of the ledger honest. Costs that a client reimburses at cost, such as a stock licence or a train ticket rebilled on the invoice, are left out, since they pass through. And the tax percentage is an average over the year, not your top bracket: using the marginal rate inflates the floor, and leaving tax out altogether sets it far too low.

How many billable hours do you really have in a year?

The billable hours in a freelance year are the working days left after time off, multiplied by the share of each day a client pays for, and the result sits far below the 2,080 hours of a 40-hour week over 52 weeks. In the worked example it is 1,056 hours.

Start from the calendar. A year holds 52 weeks of 5 weekdays, which is 260 weekdays. In the example the freelancer takes out 10 public holidays, 20 days of vacation, 5 days of sickness and 5 days of training, 40 days in all. That leaves 220 working days, and at 8 hours a day, 1,760 working hours.

Then take out the unpaid time inside those working days. It is the time a business needs to exist and it has no invoice attached:

• finding work: prospecting, calls with people who do not sign, proposals and quotes;
• running the business: invoicing, chasing payment, bookkeeping, tax filings, insurance renewals;
• keeping the work good: learning, tooling, maintaining your own site and portfolio;
• gaps: the days between two projects and the mornings lost to a client who moved a deadline.

The example assumes 60% of working hours are billable. That gives 1,056 billable hours, equal to 132 billable days, and leaves 704 hours a year for everything in the list. Over the 44 weeks actually worked, it is 24 billable hours a week. If that sounds low, measure it: a freelancer who is still building a client base may well find a lower figure.

The billable share is the one assumption worth measuring instead of guessing. Count hours invoiced over hours worked across three months, and use that figure, not the one you hope for.

What is the minimum hourly rate and day rate?

The minimum hourly rate is the revenue to bill divided by the billable hours, and the minimum day rate is the same revenue divided by the billable days. In the worked example, $116,000 over 1,056 hours gives $109.85 an hour, and $116,000 over 132 days gives $878.79 a day.

Round up, never down: $110 an hour and $880 a day. At those rates a full year of 1,056 hours or 132 days bills $116,160, which clears the target by $160. This is a floor, the price at which the year exactly works if every assumption holds. It carries no margin for a slow quarter, an unpaid invoice or a negotiation, so the rate you quote sits above it.

Compare it with the shortcut. Someone who wants to "earn $100,000" and divides by 2,080 hours arrives at $48.08 an hour. The bottom-up rate is 2.28 times that figure. Billed for the 1,056 hours really available, $48.08 brings in about $50,769 of revenue, less than half of the $116,000 needed, before a single cost is paid.

The chart shows why the billable share deserves more attention than any cost line. The revenue target stays at $116,000 and the working year at 1,760 hours; only the share of billable hours changes.

A freelancer who bills 40% of their time needs exactly twice the rate of one who bills 80%, for the same take-home. That is why the two levers of a freelance business are the rate and the share of time that is sold, and why work that arrives without a sales effort, from referrals and repeat clients, is worth more than its invoice suggests.

Hourly, daily, per-project or value-based: which pricing model fits?

The four pricing models differ on one question: who pays when the work takes longer than expected. Under hourly and daily pricing the client carries that risk; under per-project and value-based pricing you carry it, and the price has to include a premium for doing so.

The floor rate is the same under all four. What changes is the unit the client buys and what the client can compare.

Hourly pricing is the simplest to start and to defend: time is logged, the invoice follows. Its limits are equally plain: the client can compare the number with a salary, income is capped by hours, and getting faster earns you less. A job you finish in 90 hours bills $9,900 at $110 an hour, while a slower colleague bills more for the same result.

Daily pricing sells time in a larger unit and removes the argument about every quarter of an hour. Twelve days at $880 is $10,560, a figure a client can budget. It needs one definition written into the agreement: how long a day is, and whether a half day exists. Without it, a "day" stretches to ten hours and the effective hourly rate falls with no change to the invoice.

Per-project pricing gives the client what most of them want, a number that will not move, and gives you the upside of your own efficiency: a fixed price of $14,800 delivered in 90 hours works out at $164.44 an hour. The same structure works against you when the scope was loose or the estimate was thin, which is why a project price is only as good as the estimate and the scope description behind it.

Value-based pricing sets the price from what the result is worth to the client instead of from the hours it takes; it has its own section below.

Whether to turn recurring work into a monthly retainer is a separate decision, covered in the article that compares retainer and project pricing. A retainer trades a lower rate for predictable revenue and a higher billable share, and it only pays if the monthly scope is capped in writing.

How do you estimate a project price?

You estimate a project price by listing every task with its hours, adding a contingency for what you cannot see yet, multiplying the total by your hourly rate and adding outside costs. The estimate is done at task level, since a single number for "the whole thing" is where hours go missing.

The worked example is a website redesign for a small firm, estimated at the floor rate of $110 an hour so that the arithmetic stays visible. A real quote would use your market rate, which sits above the floor.

The contingency is not padding and is not shown as a line the client can strike out. It is the price of carrying the overrun risk that a fixed fee moves onto you. If the project lands on its 112 task hours, the $14,800 fee works out at $132.14 an hour and the contingency becomes margin. If it uses all 134.4 hours, you earn $110.12 an hour, your floor. Either outcome is acceptable, which is the point of the method.

A contingency does not replace a scope description. It absorbs the unknowns inside the agreed work: a page that needs a third layout, a content delay. It does not absorb new work. A request that adds a deliverable is priced as a change, at the rate named in the quote, and how to keep that boundary is the subject of the article on scope creep.

When should you quote a range instead of a fixed price?

Quote a range when the scope is still open at the moment the client wants a number, and a fixed price when the scope is written down. A range is an honest answer to an unfinished brief; a fixed price on an unfinished brief is a guess that you alone will pay for.

Three situations call for a range: a first conversation in which the client asks "roughly how much", a project whose size depends on something nobody has examined yet (the state of an existing codebase, the quantity of content to migrate), and work where the client has not chosen between options.

A range should be built, not improvised. Give three estimates for the same redesign: an optimistic 96 hours, a likely 112 and a pessimistic 150. At $110 an hour those are $10,560, $12,320 and $16,500. A workable range runs from the likely case to the pessimistic one, so $12,300 to $16,500, and each end is tied to a stated condition: the low end if content arrives final and one person signs off, the high end if the existing site has to be rebuilt instead of adapted.

Do not quote from the optimistic figure. Clients tend to budget to the low end of a range, so the bottom of yours has to be a price you would accept. For the same reason, keep the range narrow enough to be useful: a spread from $5,000 to $25,000 tells the client that you have not understood the job yet, and the right response at that stage is a paid discovery phase with its own small fixed price, after which you can quote the rest firmly.

Once the open questions are answered, replace the range with a fixed price or with a time-based engagement that has a stated cap.

How do you set a value-based price?

A value-based price is set between two limits: your cost of delivering the work, which is the floor, and the value of the result to the client, which is the ceiling. The price is negotiated inside that band, as a share of the value.

Suppose the redesign in the example is expected by the client to bring in $120,000 of additional gross profit in its first year. That estimate has to come from the client's own numbers, not from yours. A fee set at 10% to 20% of that first-year value is $12,000 to $24,000; those percentages are illustrative, since no standard share exists. At 15% the fee is $18,000, and over the 134.4 budgeted hours that is $133.93 an hour.

The ledger still does two jobs here. It tells you when to walk away: the low end of the band, $12,000, is below the $14,800 cost-based price, so any agreement under that level means working beneath your floor for the sake of the label. And it tells you what you earn: converting the fee back into an effective hourly rate after the project is the only way to know whether value pricing is paying better than time pricing for you.

Value pricing fits only some work. It needs a result that can be measured, a client who shares the numbers, and a link between your work and that result that both sides accept. Maintenance work, whose value is the absence of problems, is hard to price this way; where the value cannot be stated, a well-estimated fixed price is the sounder structure.

Which mistakes lower your effective rate?

The effective rate is what you were paid divided by every hour you actually spent, and four habits lower it without touching the rate on the quote: estimating without contingency, absorbing extra hours, discounting to close, and leaving unbilled time out of the count.

Follow the $14,800 redesign through a project that goes the way many do. The hours run to the pessimistic 150 because revisions were not capped. The client asks for 10% off to sign quickly, which brings the fee to $13,320. Then 8 more hours go into calls after delivery and into chasing the final payment.

The last line is 23.3% below the $109.85 floor. To be paid the floor rate for 158 hours, the fee would have needed to be about $4,036 higher.

The same arithmetic applies to a whole year. A 10% discount on an hourly rate of $110 leaves $99, and billing $116,000 at $99 takes about 1,172 hours instead of 1,056: roughly 116 extra billable hours, close to three more weeks of full-time work, to end the year in the same place.

Other habits do the same damage more quietly:

• quoting an hourly figure on the spot, before the scope is known, and being held to it;
• letting a day rate cover days of any length;
• forgetting payment fees and currency conversion on incoming payments, which reduce what a fee is worth after it is agreed;
• never recomputing the ledger when costs, tax or the billable share change.

The remedy for each is a number written before the work starts: a cap on revisions, a definition of a day, a rate for extra work, and a discount given only in exchange for something, such as a smaller scope or payment in advance.

How does your minimum rate relate to the market rate?

Your minimum rate tells you what you cannot go below; the market tells you what you can charge. The floor comes from your own costs and time, the market rate comes from what clients pay for comparable work, and the price you quote should be the higher of the two.

When the market rate for your work is above your floor, the gap is your margin, and there is no reason to quote the floor because you computed it. When the market rate is below your floor, the ledger is telling you something a competitor's price list cannot: at that price the year does not work. The choices are then to change the inputs (a different kind of client, a narrower specialty that commands more, a higher billable share, lower costs) or to accept a lower take-home knowingly. Matching a price you cannot afford is not one of them.

Recompute the ledger at least once a year and whenever an input moves: a new insurance premium, a change in tax treatment, a measured billable share that differs from the assumed one. When the floor rises, or when demand for your time exceeds the hours you have, your rates should follow; how to announce an increase to existing clients, and how much to raise at once, is covered in the article on raising freelance rates.