A first version of a web product that people pay for usually costs $30,000–70,000 with an experienced freelancer, and takes about 6–14 weeks. A clickable demo costs $3,000–12,000, an internal tool $15,000–35,000, and a full platform $80,000–200,000. Five products I built cost $25,000–40,000 each.
A web product is anything people log into and come back to: a SaaS tool, a booking platform, a customer portal, an internal system. The price depends mostly on how many kinds of users it serves and how much has to work on day one, not on how many screens it has. If you're not sure whether you need a product or a website, start with website or web product. Below: prices by type, five real products, what a quote should include, what drives the cost, running costs after launch, and how to spend less.
Web product and SaaS cost by type
Freelance prices in US dollars, for building from scratch. In my experience, agencies often quote about twice these numbers for the same work.
| Type | Cost | What you get |
|---|---|---|
| Clickable prototype or demo | $3,000–12,000 | Enough to pitch or test the idea. Not built for real users. |
| Internal tool or admin platform | $15,000–35,000 | Staff complete a real job in it every day. |
| First version people pay for | $30,000–70,000 | One type of customer, one core job, logins, payments, and a live product. |
| Full platform | $80,000–200,000 | Several user types, admin tools, and many integrations. |
| Type | Typical range | Real projects |
|---|---|---|
| Clickable prototype or demo | $3,000–12,000 | — |
| Internal tool or admin platform | $15,000–35,000 | — |
| First version people pay for | $30,000–70,000 | — |
| Full platform | $80,000–200,000 | — |
| Five products I built | — | Digital business card: $25,000; Printing platform: $30,000; School platform: $32,000; Personalization engine: $35,000; Farm rental platform: $40,000 |
Five real web products and what they cost
Ranges only go so far. These are five products I built, with the price, the time it took, and what happened after launch.
| Product | Cost and time | Result |
|---|---|---|
| Digital business card platform | $25,000, 8 weeks | 30K+ users, around 10K cards sold |
| On-demand printing platform | $30,000, 20 weeks | 400+ orders a day across 40+ branches |
| School management platform | $32,000, 10 weeks | 1,000+ students and 90+ staff in one system |
| Personalization engine | $35,000, 8 weeks | 100K+ personalized pages generated |
| Farm rental platform | $40,000, 14 weeks | 200+ customers and 1,000+ deliveries |
A school platform: $32,000 in 10 weeks
Hongirana is a school that ran registrations, attendance, and fees across registers and spreadsheets. I built one system for students, staff, fees, report cards, and parent messages by SMS, email, and WhatsApp, plus the public website. It now holds 1,000+ students and 90+ staff, and about an hour of daily admin is automatic. It cost $32,000 because it served several user types from day one: staff, teachers, and parents.
What a first-version quote should include
Two quotes for "the same product" can differ by $40,000 because one includes things the other quietly leaves out. A complete first-version quote covers:
- A written brief: who uses it, and the one job it must do
- Design of the key screens, on phone and desktop
- Sign-up, login, and password reset
- The core feature, working end to end
- Payments, if customers pay inside the product
- An admin view to manage users and data
- Emails or notifications the product sends
- Hosting, a domain, and a launch without downtime
- Testing, including what happens when things fail
- Handover: access, notes, and accounts in your name
If a quote is missing several of these, it's probably priced as a demo. Ask what's excluded in writing before you compare numbers. The cheapest-looking quote is often the one with the longest list of exclusions, and each exclusion comes back later as a separate invoice. Mine include all ten.
What makes a web product cost more or less
These move the price most, roughly in order. Screen count matters far less than people expect.
How many kinds of users
A product for one type of customer is much cheaper than one with customers, staff, partners, and admins, each with their own screens and permissions. The school platform above cost more partly for this reason.
Money moving through it
Payments, subscriptions, refunds, wallets, and payouts need careful handling and testing. A product that takes one payment is simpler than one that holds balances or pays people out.
Integrations and AI
Connecting to CRMs, calendars, maps, or AI models adds work for each one, especially when they must stay in sync. AI features are priced separately in what an AI feature costs.
What makes it cheaper
One user type, one core job, plain screens built from a good component library, and existing services for logins, payments, and email instead of building them.
Running costs after a web product launches
The build is the biggest cost, but not the only one. Plan for these from the first month:
| Cost | What to expect |
|---|---|
| Hosting and database | Small at launch, growing with users |
| Payment processing | A percentage of each payment, set by your provider |
| Email, SMS, and AI usage | Billed per message or request, so it grows with use |
| Support and small changes | With me, a care plan at $400, $900, or $1,800 a month |
All of these should be billed to accounts in your name. What support covers, and when paying per change is better, is in what maintenance costs.
What a cheap SaaS quote usually leaves out
A low quote isn't always wrong. Sometimes it's an honest price for something smaller than you asked for. These are the most common gaps:
Each one can be the right choice, if it's what you meant to buy. A demo is a fine way to test an idea. The problem is paying for a demo while believing you bought a product. Ask every developer the same question: if a customer signs up and pays on launch day, what happens next, and who fixes it if something breaks? If you've already built on a no-code tool, see when to leave no-code for custom.
How to lower the cost of a first version
The most reliable way to spend less is to build less, on purpose. These cut the most without hurting what users see:
- One type of user. Launch for customers first. Staff and partner tools can wait.
- One core job. The thing people would pay for, working completely. Everything else is later.
- Existing services. Ready-made login, payments, and email rather than custom versions.
- Plain design. Clean, consistent screens. Polish after real users show what matters.
- Phases. Launch the core, then let revenue fund the next part.
How to decide what makes the cut is in what to put in a first version, and how long each size takes is in how long a first version takes. If you already have an AI-built demo, it can be a useful head start, as it was for The Chat Company. Every feature you leave out of version one is money you can spend on the features real users actually ask for.
A brief to get an accurate product quote
Send this to every developer you're considering. It turns a vague "how much for an app like X?" into quotes you can actually compare:
A developer who reads this and suggests cutting something is usually one worth talking to. Someone who quotes everything exactly as asked, without a single question, usually hasn't thought about it much. Prices for this kind of work are also on my SaaS and web products page. Or send me the brief, and I'll tell you which row you're in and what I'd build first. If the honest answer is a smaller first version, I'll say so.
