Cost and budget · 10 min read
Why website quotes vary so much, and how to read the one in front of you
No price ranges here either. Instead: what actually drives the cost of a website build, why apparently identical briefs come back at wildly different prices, and the ongoing costs that get left out of almost every quote until they arrive.
Written by Zubair Afzal, FounderUpdated
The thing you came here for
Why three quotes for one brief came back three very different prices
A website quote is a bet on how much of the ambiguity in your brief will turn out to be expensive. When ten suppliers read the same document and price ten different scopes, that is usually ten different bets rather than nine wrong numbers.
These are the four ambiguities that account for most of the spread.
- Nobody agreed who writes the content.
- One supplier assumed you provide finished copy for every page. Another priced a writer, an interview process and two rounds of amends. That single assumption can be a large fraction of the whole project, and it is invisible in a headline number.
- Design origination was read differently.
- Adapting a theme to your brand and originating a design system from scratch are different disciplines with very different day counts. Both can legitimately be described as "custom design" in a proposal, and the word tells you almost nothing.
- Migration and redirects were left out of one quote.
- Moving existing content, mapping old URLs to new ones and preserving search visibility is real work that a replatform cannot skip. When it is missing from a quote it does not mean it is unnecessary — it means it will be a change request later, or it will be skipped and cost you traffic.
- Revision rounds and stakeholder count were unstated.
- A build approved by one decision-maker and a build approved by a committee of eight are different projects. Approval cycles consume time that has to be priced somewhere, and where they are unstated, the optimistic supplier looks cheaper.
The variables
What genuinely drives a website build cost
Score your own project against these. Most of the difference between a modest build and an expensive one is concentrated in the first four.
- Number of unique templates and components — not page count. Each unique layout needs designing, building, making responsive, testing and documenting.
- Design origination. Configuring an existing theme, adapting a design system, or originating one from nothing are three different jobs with very different effort profiles.
- Content production. Who writes, who photographs, who supplies product data, and how many approval rounds sit between draft and publication.
- Integrations. Every connected system — CRM, payments, booking, ERP, product feeds, marketing automation — adds specification, build, error handling and testing, and each behaves worse than its documentation suggests.
- Ecommerce complexity. Catalogue size, variants, pricing rules, tax and shipping logic, and how much of it is bespoke rather than out of the box.
- Migration. Moving existing content, mapping redirects and preserving search visibility. Non-negotiable on a replatform and frequently omitted from quotes.
- Accessibility and compliance targets. Meeting a recognised standard properly means design decisions, build discipline and testing, not a plugin.
- Performance targets. Hitting real-world speed thresholds on mobile constrains images, fonts, scripts and third-party tags, and constraints cost time.
- Languages and regions. Each additional locale multiplies content, testing and technical configuration.
- Stakeholders and governance. Approval cycles, legal review and brand sign-off are project cost even though they produce nothing visible.
Template, adapted system or fully custom
Three genuinely different products that get described with the same words in proposals. Pick on constraints, not on prestige.
| Dimension | Template build | Adapted design system | Fully custom build |
|---|---|---|---|
| What you get | An existing theme configured with your brand, content and structure. | An established component library restyled and extended for you. | Design and front-end built from your brand and your requirements outward. |
| Where the effort goes | Content, configuration and working within the theme's constraints. | Restyling, a handful of bespoke components, and integration work. | Discovery, design system creation, build, testing and documentation. |
| Time to launch | Shortest. | Moderate. | Longest, and most sensitive to approval delays. |
| Right when | Speed matters most, needs are conventional, and your differentiation is not visual. | You need brand distinctiveness without originating everything, and your requirements are mostly standard. | The site has to do something specific, or the brand experience is a genuine competitive asset. |
| Where it goes wrong | Fighting the theme on every page once you outgrow it. | Scope creeping into custom without the budget following. | Cost and timeline expanding through indecision rather than complexity. |
| Long-run cost | Low to build, potentially high to escape. | Moderate, and usually the most economical over several years. | High to build, cheapest to extend if the system was built properly. |
The part left off the quote
The cost that starts on launch day
A website is not a purchase. It is a subscription you pay whether or not you budgeted for it, and the businesses that get the most from a build are the ones that funded the second year before they signed off the first.
The recurring costs are unglamorous and non-optional: hosting, domain and certificate renewal, CMS and plugin licences, security patching, backups and monitoring, accessibility maintenance as standards and content change, and analytics upkeep. None of these produce anything visible. All of them cause visible problems when skipped.
Then there is the content. A site that was accurate at launch drifts. Services change, prices change, team members leave, the case that was your best example is now three years old. Content maintenance is the difference between a site that earns its cost over five years and one that quietly stops representing the business after eighteen months.
When comparing quotes, ask each supplier what the first twelve months after launch cost and what that includes. The answers vary as much as the build prices do, and the difference is easier to understand.
Practical
Three things that make quotes comparable
Specify templates, not pages
List the distinct page types you need — home, service, location, article, contact, product, checkout — and ask for a price per template set. Suppliers can then price the same thing, and you can see where one has included more than another.
Name the content owner per page
Go through the sitemap and write a name next to each page for who produces the words and the images. Do it before requesting quotes. It is tedious, it takes an hour, and it prevents the single most common overrun in website projects.
Ask what is excluded
Exclusions tell you more than inclusions. Migration, redirects, content, photography, third-party licences, training, post-launch support and accessibility testing are the usual omissions. A supplier who lists them clearly is easier to trust than one whose quote is a single number.
Questions
Website cost questions people ask
Why do website quotes vary so much for the same brief?
Because the brief is rarely as specific as it feels. Every supplier prices its own assumptions about content, revision rounds, integrations, migration, testing and post-launch support, and those assumptions differ enormously.
The cheapest quote is usually the one that assumed the least. That is not dishonesty; it is a different reading of an ambiguous document. The fix is to specify the ambiguous parts before asking for prices.
Is a template site ever the right choice?
Often, yes. If your differentiation is not visual, your content needs are modest and you need to be live quickly, a well-chosen template built properly will serve you better than a custom build you cannot afford to finish.
A template stops being right when your site has to do something specific — an unusual booking flow, a complex catalogue, a distinctive brand experience — or when you have outgrown the constraints of the theme and are fighting it on every page.
What is the biggest hidden cost in a website project?
Content. Words, photography, product data and the approvals around them. Both parties routinely assume the other is handling it, and the discovery arrives after the design is signed off, when the timeline is already committed.
Decide in writing who writes each page, who supplies imagery, and how many rounds of approval exist. That single conversation prevents most website overruns.
Does page count determine the price?
Much less than people expect. The real driver is the number of unique templates and components, because each one has to be designed, built, made responsive, tested and documented.
A large site built from a well-designed set of templates can be cheaper than a small site where every page is bespoke. When comparing quotes, ask how many unique templates are included rather than how many pages.
What should I budget for after launch?
Hosting, domain renewal, CMS and plugin licences, security updates, backups, monitoring, accessibility maintenance and content updates. None of these are optional, and all of them recur.
Treat a website as a subscription rather than a purchase. Sites do not fail because they were built badly nearly as often as they fail because nobody funded them after launch.
How do I make quotes comparable?
Write a brief that specifies the things suppliers otherwise have to guess: number of unique templates, who produces content, which systems must be integrated, how many revision rounds, whether migration and redirects are included, and what post-launch support means.
Then ask every supplier to price that same document. Remaining differences will be about approach and seniority, which is a useful comparison rather than a confusing one.
Have a quote you would like a second opinion on?
Send it over. We will tell you what looks missing and which assumptions are likely to become change requests later, whether or not you end up building with us.
Last updated