What a website really costs in 2026 (and what drives the number)
In short
A website's cost follows scope, content, integrations and migration, not page count. Design and build are often the smaller half; content, data migration and third-party connections are where budgets move. Ask for a fixed scope, a phased plan and a quote after discovery, and pay for the work you will actually use.
Two agencies can quote for the same brief and land an order of magnitude apart without either being dishonest. The difference is scope: what is being built, who writes the words, what it connects to, and what happens to the old site.
What actually drives the number
A website is priced like a building: by what it has to hold. Two briefs can look identical in a proposal and differ by a factor of five in scope, and the difference hides in the detail.
The drivers repeat across projects. Scope, content, integrations, migration, languages, and the standard you set for performance and accessibility.
Ask any agency to explain which of these is driving your quote. If they cannot, the quote is a guess.
The number people remember is the build. The number that decides the budget is everything around it: content, data, connections, and the plan for what happens after launch.
- Scope: how many templates, how custom the design, how many states and edge cases.
- Content: who writes, shoots, translates and structures everything.
- Integrations: CRM, ERP, payment, booking, search, email and the systems around them.
- Migration: how much existing content, data and redirect history has to survive.
- Standards: performance, accessibility, SEO and legal requirements you commit to.
Read budget bands, not price lists
Serious agencies rarely publish price lists, because scope moves. What they can publish is how they think in bands.
A focused marketing site with a handful of templates sits in the lowest band. A content-heavy site with custom design, a CMS, several languages and analytics sits in the middle. Commerce, personalisation, logins and deep integrations sit above that.
Clika quotes sur devis after a scoping session, because the useful number only exists once your content, data and connections are understood. Before that, a fixed price is either padded or fiction.
The practical move is to decide the band before you compare agencies. Then compare the approach inside the band: who does the strategy, who writes, who owns the result.
Content is half the budget, and the half people forget
Content is the most common reason a website runs late, and the most common reason a second budget appears halfway through. Copy, photography, translation and product data are not small extras; they are the site.
The cheapest content strategy is to decide early. A team that arrives with approved messaging, a photography plan and a page-by-page structure saves real money in the build.
If you cannot supply the content, buy it deliberately. One writer and one photographer inside the project will cost less than a year of patching empty pages after launch.
Price the content in the same conversation as the build. If it is missing from the proposal, it has not disappeared; it has moved to you, unfunded and unscheduled.
How to avoid paying twice
Most brands pay twice because the first scope was wrong, not because the first agency was bad. They buy a design, then a rebuild, then a migration to a platform that fits.
Paying again can be rational when the scope genuinely changed. Paying again for the same work is avoidable, and it is usually a governance problem rather than a technical one.
Put the exit terms in the contract: what you receive, in what format, and when. A website you cannot take elsewhere is a subscription, not an asset.
- Decide platform and structure before design, not after.
- Keep one source of truth for content, outside the website.
- Own the domain, the hosting, the repository and the analytics account from day one.
- Cut features nobody will use; a smaller launch is faster to fix.
- Plan the second phase before the first one ships.
What you should own when the project ends
At the end of a project, the list of what you own matters more than the launch. Ask for it in writing before you sign.
You should hold the domain, hosting and CMS accounts under your own emails, the source code, the design files, the content and the analytics history. Access should survive any agency change.
Then put one person in charge internally. The biggest risk to a new site is not the build; it is nobody owning it afterwards.
A short handover session, recorded and documented, is worth more than an extra month of support you never call on. Support is not a substitute for understanding your own site.
- Accounts you control: domain, hosting, CMS, analytics, email.
- Files you hold: source code, design files, image library, copy documents.
- Documents you keep: redirect map, schema, measurement plan, training notes.
- A named internal owner with the access and the time to use it.
Because scope decides the number. Ask for the drivers, the band you are in and what would move you between bands. That tells you more than a price list ever will.
Usually cheaper to start and more expensive to grow. Templates are fine when your needs are standard. The costs appear later in limits, licensing and the rebuild you did not plan.
One honest scope, ready content and a single decision-maker. The band is lower than most brands fear if they arrive with those three things.
A focused marketing site is a matter of weeks; commerce with integrations and a migration runs into months. The clock starts when content decisions are made, not when the contract is signed.