One of these billing approaches quietly punishes you for getting good at your job. After 25 years of billing both ways, here's how I decide which to use - and what neither of them can fix.
Ask ten freelancers whether to bill by the day or by the hour and you'll get ten confident answers, most of them based on whatever the person happened to do first. It's a decision worth more thought than that, because the billing unit you choose quietly shapes how clients treat your time, how much of your work gets paid for, and how hard it's to raise your rates later.
I've billed both ways over 25 years. Here's how I decide.
Charge a day rate when the work needs focused blocks of your attention and the client is buying a chunk of progress: sprints, on-site engagements, design and build work, senior advisory retainers. Charge hourly when the work arrives in unpredictable slices you can't batch: ongoing maintenance, small fixes, ad-hoc support, and anything with genuinely open scope. The day rate protects your margin, because it stops the client from auditing your minutes and stops you from being punished for working fast. The hourly rate protects the client, which is why it survives in support work where nobody can forecast the volume. Both are still time-based billing, and both have the same ceiling, so treat whichever you pick as a stepping stone toward pricing the outcome rather than the clock.
Mechanically, one bills in units of an hour and the other in units of a working day, usually defined as somewhere between six and eight hours. The important difference is behavioral.
An hourly rate invites supervision. The client is buying minutes, so the client tends to watch minutes: time trackers, screenshots, questions about why a task took three hours instead of two. It also creates the efficiency penalty that every experienced freelancer eventually notices. The better you get, the faster you finish, and the less you earn for the same job. Twenty-five years of practice made me roughly three times faster at certain work. Billing hourly, that expertise would have cut my income by two-thirds.
Hourly billing pays you for being slow and penalizes you for being good. That is a strange incentive to build a career on.
A day rate moves the conversation up a level. The client buys your Tuesday, and what happens inside that Tuesday is your professional business. Nobody asks whether you took forty minutes for lunch. It also makes scheduling honest: a day is a real, countable thing on a calendar, so both sides can plan around it, and you can hold a client to booking days in advance rather than expecting you to appear on demand.
Day rates fit work that benefits from immersion and comes in identifiable chunks. Four situations where I reach for one:
Focused project work. Design, development, writing, strategy - anything where context-switching destroys quality and a half-hour here and there produces nothing worth billing for.
On-site or synchronous engagements. If you have to be somewhere, or be available for a whole working day, you're giving up the entire day regardless of how many hours the work technically takes.
Sprint-shaped delivery. "Three days to migrate the booking system" is easy for a client to approve and easy for you to schedule. It also caps your exposure to scope creep, since day four is a new purchase.
Senior or specialist work. The more your value comes from judgment rather than labor, the worse hourly serves you. Expertise compresses time, and the day rate is the first billing model that stops punishing you for it.
Day rates also raise your effective floor without a fight. Clients who would balk at "$150 an hour" will book a "$1,200 day" without blinking, and the arithmetic is nearly identical. The unit changes what the number is compared against.
Hourly isn't obsolete, and pretending otherwise would be dishonest. It earns its place when the work genuinely can't be batched into days.
Ongoing maintenance and support. Twenty minutes on Monday, an hour on Thursday. Nobody can forecast that, and pretending it's a day of work isn't fair to the client.
Truly open-ended scope. When neither of you can describe the finished state yet, an hourly meter is the honest instrument while you figure it out.
Advisory and calls. A 90-minute strategy session priced by the hour is clean and easy to say yes to.
Early career. If your estimates are still shaky, hourly is a reasonable place to start while you learn how long your own work actually takes. Just plan your exit from it.
If you stay hourly, set a minimum increment - a one-hour or two-hour minimum per request - so that a five-minute fix doesn't silently become unpaid work. That single clause is worth more than most rate increases.
Day rate | Hourly rate | |
|---|---|---|
Best for | Focused project blocks, sprints, on-site work, senior advisory | Maintenance, support, ad-hoc fixes, open scope, calls |
Who it protects | You. Speed and expertise stop working against you. | The client. They pay only for time actually spent. |
Scope creep | Contained. More work means booking another day. | Unbounded, but metered - the bill just keeps running. |
The failure mode | A "day" quietly stretching to ten hours because it was never defined. | Getting audited by the minute and penalized for being fast. |
Not by multiplying your hourly rate by eight. That math bakes in a fantasy: that you bill eight productive hours a day, every working day of the year.
You don't. Proposals, invoicing, client calls, marketing, bookkeeping, software, taxes, sick days, holidays, and the gaps between projects are all real and none of them are billable. A freelancer who works hard and manages their pipeline well might sell 120 to 150 fully billable days in a year. Build your rate on 250 and you have designed a business that loses money at full capacity.
The honest sequence runs the other way:
Start with your cost of doing business. Everything an employer used to cover and now doesn't: health insurance, retirement contributions, equipment, software, professional development, liability insurance, and the 15.3% self-employment tax on your net earnings (applied to the first $168,600 for 2024; the cap moves most years).
Add the income you actually need to live, plus a real profit margin on top. Profit is what pays for slow quarters and the year you take a proper vacation.
Divide by the billable days you can realistically sell, not the days on the calendar.
The number that falls out is your floor - the rate below which you're working at a loss while feeling busy. Whatever you charge above it's a pricing decision. Whatever you charge below it's a slow-motion problem.
This is the exact math Pricing Compass was built to do. It walks you through living expenses, self-employment costs, taxes, savings, and realistic work capacity, then tells you the rate your actual life requires. It's free to start, and it's the only calculator I know of that includes healthcare, retirement, insurance, and taxes rather than the naive "income plus expenses divided by hours" arithmetic.
The reason this matters more than the day-versus-hour question: most freelancers set a rate by looking at what other freelancers charge, then discover the shortfall years later. Pew's research found 41.6% of nontraditional workers have no defined-contribution retirement savings at all, and among sole proprietors it's 54.7%. That's not a discipline problem. That's a pricing problem, compounding quietly.
The most common way a day rate goes wrong is that nobody wrote down what a day is. Put it in the agreement, in plain language:
The length of a working day (I use seven hours, and I would rather define it as seven than pretend it's eight).
What happens past that. Overtime at a stated rate, or it waits until tomorrow.
Half-days, if you offer them, and their price.
Cancellation terms. A booked day inside 48 hours is billable. You turned away other work to hold it.
What the day includes. Meetings and reviews count as work, and they belong inside the booked time rather than as a free extra.
An undefined day always expands. It never contracts.
Here's the thing: day rates and hourly rates are the same product in different packaging. Both sell time. Both cap your income at hours-in-a-life, and both leave the money on the table when your work is worth far more to the client than it cost you to do.
When you fix a booking system that loses a studio four appointments a week, the value of that fix has nothing to do with whether it took you six hours or two days. Pricing it by the clock is a decision to hand the client the difference. That's why the endgame is pricing the outcome - a project fee, a productized service, a retainer priced on what it protects rather than what it consumes.
My advice for most working freelancers: move from hourly to day rates as soon as your estimates are reliable enough to hold, and use day rates as the bridge to fixed-price offers built around a specific, expensive problem you solve for a market you chose deliberately. That progression is the whole game. It starts with knowing your floor.
Default to a day rate for project work, keep an hourly rate (with a minimum increment) for support and open-ended work, and price the outcome wherever you can see the client's payoff clearly. Write the definition of a day into every agreement. And before any of it, run your real numbers so you know what your floor is.
Do that math first with Pricing Compass. Once you know what your life actually costs to run, the day-versus-hourly question stops being a debate and becomes arithmetic. If you want the fuller picture on what the platforms pay for the same work, the numbers are assembled in Are Upwork and Fiverr Worth It?, and the benefits you now have to buy yourself are broken down in how to replace your corporate benefits.
Good luck. Charge what the work is worth.
For focused project work, yes. A day rate stops the client from auditing your minutes and stops your speed from cutting your income, which is the built-in penalty of hourly billing: the better you get, the faster you finish, and the less you earn for the same job. Hourly still makes sense for maintenance, ad-hoc support, advisory calls, and genuinely open-ended scope, where the work can't be batched into whole days.
Don't multiply your hourly rate by eight. Start from your cost of doing business (health insurance, retirement, equipment, software, insurance, and the 15.3% self-employment tax), add the income you need to live plus a real profit margin, then divide by the billable days you can realistically sell in a year - typically 120 to 150, not 250. The result is your floor; anything you charge below it means losing money at full capacity.
Usually six to eight, and the number matters far less than writing it down. Define the length of the working day in the agreement, state what happens past it (overtime at a stated rate, or it waits until tomorrow), price half-days if you offer them, and set cancellation terms for booked days. An undefined day always expands.
Hourly is a reasonable starting point while your estimates are still shaky and you're learning how long your own work really takes. Set a one-hour or two-hour minimum per request so small fixes don't become unpaid work, and plan your exit: move to day rates as soon as your estimates hold, then toward fixed-price offers built around a specific problem you solve.
Put this into practice with the Freelance Mentors toolkits.
This article lives in The Freelance Plan, the preparation stage of the expedition.
More guides and free toolkits to prepare for the move: readiness, runway, niche, and your plan.
Return to the hub
The full guided ascent: 120+ lessons, the complete toolkits, and mentorship from first client to thriving business.
Explore the program
My free Saturday morning newsletter delivers freelance wisdom and advice straight from my own personal experience and research.
It’s short and sweet but nutritionally dense - the perfect complement to your weekend morning coffee.
No fluff, no spam. Unsubscribe anytime.
