A client emails asking you to quote something you've genuinely never done. Not a variation on your usual work — something new. Your first instinct is to throw out a number just to keep the conversation moving. Don't. That instinct is how freelancers end up trapped in 40-hour "quick jobs" — and it's exactly the situation where knowing how to price a project you've never done before actually pays off.
How to Price a Project You've Never Done Before: The Short Version
Say you want to give the client a number you can actually stand behind, not a guess, and ask for a day or two to look at the specifics before you name one. Then use that window to split the project into the parts you already know how to price and the part you don't. Pad or structure your quote around that unfamiliar piece specifically, instead of guessing at one blended number for the whole thing. Create the estimate or invoice once you've landed on a number → — free PDF generator, no signup.
Don't Quote in the Room — Buy Yourself Time First
The single biggest mistake with unfamiliar work is answering "how much would this cost?" on the spot. You don't have the information yet, so any number you say is a guess dressed up as a quote.
Instead, say something like: "I want to give you a number I can actually stand behind, not a guess — let me look at the specifics and get back to you by [day]." That's a completely normal thing to say, and clients respect it more than a number that turns out to be wrong.
Use that window to ask the questions that actually move the price:
- What does "done" look like? Get the client to describe the finished thing, not just the general idea.
- Are there examples of similar work — theirs or a competitor's — you can look at?
- What's the timeline, and is it flexible?
- Is there a budget range already in their head? Ask directly. Plenty of clients will tell you if you ask, and it saves both sides from wasted back-and-forth.
Split the Project Into Knowns and Unknowns
Once you have answers, break the work into two piles instead of treating it as one big mystery.
Knowns — pieces that overlap with work you've already priced before. If a chunk of this project resembles something you've delivered, you already have real data on how long it takes. Price that part with your normal rate, no padding needed.
Unknowns — the genuinely new pieces. In practice, this is usually a smaller slice of the project than it feels like at first glance. Sit down and name the parts, and most "brand new" projects turn out to be mostly familiar mechanics wrapped around a smaller unfamiliar core. Naming that core specifically — instead of treating the whole project as one undifferentiated unknown — is what makes the next step possible.
Pick the Right Pricing Model for the Uncertainty You're Carrying
Once you can see the knowns and unknowns separately, match the pricing model to how much uncertainty is actually left.
Mostly known, small unfamiliar piece → Quote a flat price, but pad the unfamiliar piece specifically:
Padded Estimate = Base Time Estimate × (1 + Uncertainty Buffer)
- 10-15% buffer for work that's new but adjacent to something you've done
- 25-35% buffer for work where you're guessing at the approach
- 50%+ buffer (or don't quote flat at all) when you genuinely can't picture the steps yet
These ranges are a working rule of thumb from how freelancers commonly scope unfamiliar work, not a published statistic. Adjust them up or down as you build your own track record of estimate-vs-actual on new work.
Worked example: you estimate a project at 20 hours, and one unfamiliar piece makes up roughly a third of it. You're fairly confident on the familiar two-thirds but genuinely unsure on the rest. So you apply a 30% buffer to the whole estimate rather than just the unfamiliar slice. It's simpler to communicate, and it covers the fact that unfamiliar work tends to create small friction in the "known" parts too (more questions, more revisions).
20 hours × 1.30 = 26 hours, billed at your normal hourly rate.
Genuinely unfamiliar, can't estimate hours with any confidence → Don't quote a flat fee at all yet. Sell a short, paid discovery phase instead — a capped block of hourly time, where your only deliverable is a real scope and a real quote. A couple of hours up to half a day is a reasonable starting range; adjust it to how big the unknown actually is. This does two things: it gets you paid for the research you'd otherwise be giving away for free, and it means your eventual flat quote is built on actual knowledge instead of a guess. The estimate maker is a clean way to send that discovery-phase proposal separately from the main project quote.
Quote a Range, Not a Single Number
When a client asks for a ballpark before you've done any research, resist the urge to name one specific figure. A range does the same job without locking you into a number you might regret:
"Projects like this typically run in the low-to-mid thousands depending on exactly what we land on for scope — I'll have a firm number for you after we talk through the details."
The width of the range should reflect your actual uncertainty (the wording above is a script to adapt to your own rates and typical project size, not a number to copy literally). A narrow range signals confidence; a wide one is honest when the unknowns are real. Either is better than a single number said too early. Walking a number down after quoting it high reads as generous — walking it up after quoting it low reads as a bait-and-switch, even when it isn't one.
The Three Approaches, Side by Side
| Flat Quote + Buffer | Paid Discovery Phase | Quoted Range | |
|---|---|---|---|
| Best for | Mostly-familiar project, one unfamiliar piece | Can't estimate hours at all yet | Client wants a ballpark before any research |
| You commit to | One padded number, upfront | A small, capped block of hourly time first | A wide window, narrowed later |
| Client gets | Price certainty immediately | A real quote, but only after paying for the scoping | Enough to plan, not enough to hold you to |
| Your risk | Low if the buffer was sized correctly | Very low — you're paid to learn the unknowns | Low — nothing's promised yet |
These aren't mutually exclusive. A genuinely unfamiliar project often starts as a quoted range in the first email, then becomes a paid discovery phase once the client's serious. It ends as a flat quote with a buffer once that discovery phase tells you what you're actually dealing with.
Break It Into Milestones to Cap Your Downside
Even a well-padded estimate can still be wrong on genuinely new work. Milestone billing limits how wrong it can get. Instead of one number for the whole project, you price and bill in phases, with a checkpoint between each one to confirm you're still on track before committing to the next chunk of work.
If the first milestone runs long, you find out early — while there's still time to re-scope the rest of the project — instead of discovering it two weeks before the final deadline. The milestone billing guide covers how to structure the phases and the invoices that go with them.
What to Do If You Underquoted Anyway
It happens, even with padding. The unfamiliar piece turns out to be harder than either of you expected. Two things matter here:
- Flag it as soon as you notice, not at delivery. The moment you can see the project is running past the estimate, tell the client — with specifics, not a vague "this is taking longer than planned." Show them exactly what's eating the extra time.
- Don't eat the whole overage silently and don't spring a surprise final invoice either. If the scope genuinely grew (not just your original estimate being off), that's a change order, and the scope-creep invoicing guide covers how to bill for it fairly. If your estimate was simply wrong on work that didn't change scope, that's a different situation. Chalk it up as a lesson, price it correctly next time — but say so to the client either way rather than letting a bloated invoice speak for itself.
After It Ships, Turn the Guess Into Data
The best thing about pricing something new is that you only have to guess once. Track your actual hours against the estimate — including the unfamiliar piece specifically — and you've turned this project into real data for the next one. The freelance rate-setting guide is worth revisiting once you have a few of these data points; your buffer percentages should shrink over time as "unfamiliar" work becomes "work I've priced before."
Quick Decision Checklist
- Is most of the project familiar with one genuinely new piece? → Flat quote with a targeted buffer on the unfamiliar part.
- Can't picture the steps at all yet? → Sell a paid discovery phase before quoting the full project.
- Client wants a number right now, before you've researched anything? → Give a range, not a single figure.
- Project has real risk of running long? → Break it into milestones so you find out early, not at the deadline.
- Still not sure after research? → Pad harder, or stay hourly for this one and revisit flat pricing once you've delivered it once.
Frequently Asked Questions
How big of a buffer should I add for work I've never done?
There's no universal number — it scales with how unfamiliar the work actually is. A small buffer (10-15%) fits work that's new but close to your normal work; a much larger one (30%+) fits work where you're genuinely guessing at the approach. When you can't estimate hours with any confidence at all, that's a sign to sell a paid discovery phase instead of padding a guess.
Should I tell the client I've never done this exact type of project before?
You don't need to volunteer it unprompted, but don't claim experience you don't have if asked directly — that's the fastest way to lose trust if it comes out later. Try framing it as "I haven't built this exact thing before, but here's the adjacent work I have delivered and how I'm approaching the new piece." Honest framing like that usually lands fine with clients who value it.
Is it OK to charge for a discovery phase separately from the project?
Yes — treat it as its own small paid engagement with its own scope (a defined number of hours, ending in a written estimate), not an unpaid sales call. Clients who balk at paying for scoping are often a preview of how they'll treat scope changes later in the project.
What if the client wants a fixed price and refuses hourly or a discovery phase?
Quote a flat number with your uncertainty buffer built in, and be transparent that the number reflects some genuine unknowns. Protect yourself with a written scope document listing what's included — so anything outside that scope is a change order, not an argument.
Put It to Work
That's the whole approach to how to price a project you've never done before: split it into what you already know how to price and what you don't. Pad the unknown piece specifically instead of guessing at a single blended number, and don't be afraid to sell a short paid discovery phase when you genuinely can't estimate hours yet. Once you've got a number you can stand behind, send a clean estimate or invoice — free, no signup, ready in under a minute.