Nobody can price a website over the phone, and almost everybody tries.
The number isn’t hiding. It’s a consequence of decisions the business usually hasn’t made yet: who writes the words, what the site has to connect to, and who needs to be able to change a page on a Tuesday afternoon without raising a ticket. Settle those three and the quotes stop looking random.
I’m not going to print bands. A range honest enough to cover everyone would be too wide to help, and a tight one would be wrong for most of the businesses reading this. What I can do is show you where the money goes, so you can read a quote properly and argue with it.
For what it’s worth, this is a question people ask constantly in meetings and rarely in a search bar. “How much does a website cost in Australia” returns no measurable volume in our Ahrefs data, while “ecommerce web design” sits at around 1,100 Australian searches a month at a difficulty of 24 (Ahrefs, August 2026). People search the category and ask the price out loud.
What you’re actually paying for
Three things arrive on one invoice, and only one of them looks like a website.
The first is decisions: what the site is for, what it says, how someone moves through it, what happens to the pages you already have. The second is production, which is design and build and testing. The third is the years afterwards, and that one is usually missing from the quote entirely, which is where most of the arguments start.
The build types, and what moves the number
| Build type | What it suits | What actually moves the price |
|---|---|---|
| Template on a hosted platform | A standard offer, a small site, speed over everything | How hard you fight the template |
| Custom design on a standard CMS | Most businesses with a sales story to tell | The number of distinct page types, not the number of pages |
| Ecommerce | Anyone taking payment online | Catalogue size, shipping and tax rules, and every system it talks to |
| Custom build or portal | Logins, quoting tools, anything with an application behind it | Integrations, edge cases and the testing they demand |
The row people misread is the second one. Ten pages that share one layout cost far less than four pages that each need their own. When you’re comparing quotes, count page types.
Where the money actually goes
Content. Every time. Words and photography are the line item that blows the timeline, because they’re the part the client owes and the part nobody has time for. A build waiting three months on copy costs real money in held resources, and “we’ll write it ourselves” has delayed more launches than any technical problem I’ve seen.
Then the connections. A CRM, an inventory system, a booking engine, a payment gateway, an ERP. Each one is an agreement between two systems that were not designed to talk, and each has behaviour nobody discovers until testing.
Then migration. Your existing URLs, the redirects that keep them working, the content that has to move. Skipping this is how a business loses a decade of search equity in an afternoon.
The design itself, honestly, is rarely the expensive part.
The cheap decisions that cost more later
Buying a site only the developer can change. Every change becomes a request, a quote and a wait, and within a year the site is out of date because updating it is annoying.
Launching without a redirect map. The rankings you paid for over ten years point at URLs that no longer exist, and the traffic drop shows up a fortnight later when nobody is watching.
Buying a design that fits the pages you have rather than the ones you’ll need. The second product line arrives and there’s nowhere to put it.
Letting the agency own the domain, the hosting account, the analytics or the code repository. This one is quiet until the relationship ends. Ask the question before you sign, and get the answer in writing.
Going live without measurement in place. If analytics and conversion tracking start the week after launch, you’ve given up the before-and-after that would have told you whether the money worked.
The costs nobody quotes
A website is a fixed asset with running costs, and it behaves like a vehicle rather than a piece of furniture.
Hosting, domains and certificates. Platform and plugin updates, which are security work, not housekeeping. Small changes, because there are always small changes. And content, because a site nobody adds to stops earning its keep in search within about a year. Budget a monthly figure for all of it before you sign, or it will come out of a marketing budget that was allocated to something else.
Two builds, two different bills
Revolve Partners came to us with decades of standing in executive search and no digital footprint at all. Marilyn & Sons did the design, we did the build: a custom WordPress architecture put together for speed, security and internal management. The money there went into architecture and into making the thing manageable by their own team rather than by us.
m|devices was a different shape entirely. An Australian medical device manufacturer with an established B2B audience opening a direct-to-consumer channel, which meant research into how consumer behaviour differs from clinical procurement, then journey and interaction design, then building it alongside their internal technology team so it fit the systems already running the business. Compliance sat over the whole thing.
Neither of those invoices was mostly design. They rarely are.
The questions to put to whoever quotes you
Who owns the domain, the hosting, the analytics and the code when we’re finished? The answer should be you, on all four.
Who writes the content, and is that in your number? If it’s you, add the weeks to the timeline honestly.
What happens to our existing URLs? You want to hear the word redirects without prompting.
What is explicitly excluded? A proposal with no exclusions is a proposal nobody prepared.
What does a change cost in month four? This tells you whether you’re buying a site or a subscription to the agency.
Who fixes it when it breaks on a Sunday night? And what does that arrangement cost.
Judging the price against the return
Give the site a job before you buy it. Enquiries a month, orders a week, applications a quarter, whatever the business runs on. Then work out what one of those is worth to you and how many the build has to produce before it has paid for itself.
That single step turns the conversation from taste into arithmetic. Without it, “$X for a website” is an aesthetic argument, and aesthetic arguments never end.
When to spend less
If you’re still testing whether anyone wants what you sell, buy a template and put the difference into finding out. A beautiful site for an offer nobody wants is an expensive way to be wrong.
If the current site is failing at one specific job, fix that job. A full rebuild is often the costly way of avoiding a decision about what the site is actually for.
And if you can’t say what the site needs to do in one sentence, that’s the work to do first, not the build. That conversation is the front end of everything in our technology practice, and the build itself lives in web development. Bring the list of systems it has to talk to. It makes the number real faster than a brief does. Tell us what you’re trying to fix and we’ll do the arithmetic with you.
