How to price web projects without guessing
Pricing is the part nobody likes to do. I spent years charging hourly like it was the only way that existed. Result? I earned way less than I should have.
After making some ugly mistakes, I discovered there’s method to it. You charge better when you stop thinking you’re “too expensive” and start understanding what you actually cost.
First step: calculate your cost per hour
This is obvious but nobody does it. Sit with a spreadsheet and calculate how much it costs to exist.
Your workspace rent (if you work from home, $20-40 monthly is fair). Internet. Computer (you’ll need to replace it every 3-4 years, amortize it). Software and tools. Taxes and contributions. Vacation and downtime (you can’t work 365 days).
I’ll use real numbers from my own situation. Fixed costs: roughly $500 monthly. Considering 4 weeks of actual work per month (I lose 2 weeks to sick days, vacation, gaps between clients), that’s 160 billable hours. Divided: $500 / 160 = $3 per hour just to cover costs.
Now you want profit. I want 50% margin. So I need to charge minimum $6 per hour. I charge more because I want breathing room to invest and grow.
Calculate yours. Should land somewhere between $8-20 per hour depending on your experience. If you’re outside that range, something’s wrong. Either you’re way overpriced or drastically undercharging.
Method 1: hourly (the easiest)
Charging hourly is simple. Client gets clarity. You get clarity. What’s the problem then?
You’re incentivized to work slowly. The slower you are, the more you earn. That’s perverse. Plus the client gets nervous watching hours pile up.
Use hourly only for work you can’t estimate. Small tweaks. Consulting. Debugging something weird. Always cap it: “first 10 hours are included, then it’s $X per hour”.
Method 2: fixed price (per project)
The opposite. You propose a price and that’s it. Client knows exactly what they’re spending.
Here you need to estimate well. If you lowball, you’re screwed. If you overshoot, you lose the client.
The formula I use: estimated hours x hourly rate + buffer.
Simple clinic website: I estimate 40 hours of work. 40 x $10 = $400. But I add 25% buffer for unexpected problems. Total: $500.
Common problem: you always estimate too optimistically. Learn from experience. If you said 40 hours and it took 60, next time you put 75. Your track record is your best teacher.
Base price for simple site: $500-800. Site with custom features: $1.000-2.000. SaaS or complex app: $4.000+.
These are my numbers. Yours will vary with experience and location. But it’s a reference point.
Method 3: value-based (where you actually make money)
Here you don’t look at your hours. You look at what the result is worth to the client.
Client with a clinic handling 50 patients monthly, each paying $100 - if your site increases conversion by 10%, that’s $500 extra revenue PER MONTH. You charging $1.000 to do that is a steal. You could charge $3.000.
This is the most economically correct method. Your gains align with client gains.
Problem: it takes guts to propose this. Client thinks it’s expensive. You need to justify the value. “Your current system converts 2%. This will convert 5%. That’s $500 extra revenue monthly.”
I started using value-based pricing in 2024. My average project price jumped 40%. And clients are happier because results matter more than hours worked.
When to charge more, when to charge less
I charge more when: client has solid budget (turned into SaaS or big company), project is urgent (rush pricing), client brings heavy complexity (scope changes daily), you’re doing something few people can do.
I charge less when: it’s a social project (NGO, association), it’s for portfolio reference, the client brings more clients, it’s small but quick for you (5 hours total).
And one important thing: you can use different pricing for different phases. Project phase: $1.000 fixed. Then need tweaks? Hourly. Need recurring maintenance? $200 monthly.
What matters is what you earn at the end. If hourly makes you more money, fine. If value-based is better, use that. The goal is to not starve and sleep well at night.
Pricing by client stage
One detail nobody mentions: new client versus returning client changes your pricing.
New client: you don’t know their workflow. There’ll be friction, alignment, extra communication. Price should include that. R$5k for a project that’s technically R$3k because you’re using goodwill to prove competence.
Returning client: you already know them. Process is tuned. Fewer checklists, less re-communicating. Can discount 10-15%. Happy client, faster new work.
Large/corporate client: rigid processes, slow approvals, late payment (90 days). Price should include that. Either raise your rate or pass on the project.
Negotiation: when and how
Your price won’t always stick. Client says: “I only have R$3k”.
Here you choose:
- Reject and stay happy
- Renegotiate scope. “R$3k? No tests. No docs. No post-deploy support. Just core. Deal?”
- Offer installments. “R$5k split three ways: kickoff, beta review, delivery.”
- Propose value-based. “Your system will make R$2k extra monthly. I’ll charge R$8k, in 4 months you’re paid back.”
Negotiation is fine. But never lower price without lowering value. That’s just selling to sell yourself short.
Clients asking for discounts
Clients always ask for discounts. “Give your buddy a discount?”
My answer: “Discount? I can reduce scope. Remove docs, remove tests, remove support. Then it’s cheaper. Want that?”
Most back off. Because they understand low price means something gets cut.
Some agree. “Just drop the docs”. Perfect. You make less but save time.
Never lower price with the same value. That’s wage slavery.
Suggested pricing by project type (2026)
| Project type | Small scope | Medium scope | Large scope |
|---|---|---|---|
| Landing page | R$2-3k | R$4-6k | R$8-12k |
| Clinic site | R$3-5k | R$6-10k | R$12-18k |
| Basic e-commerce | R$5-8k | R$10-15k | R$20-30k |
| SaaS MVP | R$15-20k | R$30-50k | R$60k+ |
| Complex web app | R$10-15k | R$25-40k | R$50k+ |
These numbers are based on junior-senior devs in Brazil, 2026. Yours can vary widely based on experience and location.
- Calculate my real cost per hour
- Review last 10 projects and adjust pricing
- Document standard estimates by project type
- Propose next client with value-based method
- Keep history of estimated vs actual time
Read also: What to charge for small tweaks | Freelance dev contract essentials | Scope is the most important dev skill
Charging well is ok. You’re offering real value. Start believing that.