A website quote looks like a price and is a description of scope dressed as a number. Two quotes can be compared only once you know what each assumes — and they usually assume different things, because they were written from the same vague enquiry.
This article is about the document and the contract behind it: which billing model to take, what to tie instalments to, and where the contingency goes. It is not about why one technology costs more than another — that is a separate matter, settled before anyone issues a quote.
Before a number appears, the supplier settles several things — and if they did not ask you, they settled them by assuming.
They need to know how many pages there are and which are unusual, because ten pages repeating one pattern is different work from three that each look different. They need to know where the content comes from — whether you supply text and photographs or they prepare them. And what the site has to connect to, because integrating with a system they have not seen is a risk that has to be priced somewhere.
And they need to know when, because a deadline is a line item, not an organisational parameter.
The conclusion runs against intuition: the more detail you give in the enquiry, the lower the quote tends to be — not because anyone concedes, but because the margin for unknowns disappears. "How much does a website cost for my company" guarantees an answer with a buffer, because anything else would be irresponsible.
If more detail means a lower quote, it is worth knowing which details change anything. No technical specification is needed — one page will do.
A list of pages by name. Not "a business site with subpages", but: home, services with three offerings, about, work, blog, contact. That list alone removes most of the divergence.
Three sites you like, and one sentence on why. The cheapest way to convey visual expectations, and the only one better than adjectives.
Who writes the copy and takes the photographs. One sentence. It settles an item that can be a tenth of the budget.
What the site has to connect to. The system's name is enough — CRM, booking, stock control. The supplier will check whether it has an API.
The deadline and where it comes from. If it is fixed by a trade fair or a campaign, say so immediately. A quote with a deadline and one without are two different quotes.
An enquiry in that form gives you offers that can be laid side by side — the only condition under which comparing amounts means anything. Without it every supplier prices their own imagination, and whoever imagined least wins.
The first decision in the document, and the one most often taken without anyone realising it is one.
Fixed price | Time and materials | Milestones | |
|---|---|---|---|
Who it suits | simple sites with a closed brief | larger sites and shops | most standard projects |
Main risk | the supplier prices in a risk margin | the budget grows unchecked | requires real sign-off of each stage |
Flexibility | none — every change is a variation | very high | moderate, within a stage |
The surprise in that table: fixed price is often the hidden expensive choice. A supplier quoting a fixed figure takes on the entire risk of unknown requirements and adds a margin for it — rightly, because the alternative is working below cost. You pay for uncertainty you introduced yourself, by giving a general brief.
With a well-described scope, hourly or staged billing often comes out cheaper, because that margin disappears. With a loose scope the reverse is true, and fixed price protects you from an open-ended bill.
The rule: the better you know what you want, the more the flexible model pays. The less you know, the more the rigid one pays — and the more that rigidity costs.
A 30/40/30 split is a convention and means nothing in itself. What means something is what triggers each instalment — an instalment with no acceptance condition is a payment date, not a milestone.
Payment instalments and what each one should accept
Digital Vantage
The middle instalment matters most, because it falls when changes are still cheap. After the wireframes are approved, every layout change means undoing work already done. So its condition should not be "half the schedule has passed" but "we have seen and approved how this is going to look".
The final instalment is the only leverage you have left after launch. Tie it not to publication alone but to the handover of access credentials and the closing of the defect list — a site that works but that you cannot change yourself has not been accepted.
A contingency of 15–20% gets treated as pessimism. It is not — and it is worth knowing what it goes on, because it is usually the same three things.
Waiting on your side. The team waits for content nobody prepared while the time is reserved. On hourly billing you pay directly; on fixed price you pay in a slipping deadline. The commonest cause of overruns and the only one you fully control.
Scope added mid-project. "Could we also add an English version?" — asked after the wireframes are approved, sounding like a detail, meaning a second version of every page, a second set of content and a rebuilt navigation. Not a bad request; a new scope, and it should be priced as one.
Licences and integrations nobody planned for. A paid plugin without which a feature does not work. A system whose API turns out different from what was assumed. A certificate somebody has to buy. Individually small, together they make up those percentage points.
What all three share: none of them is a pricing error. They are things nobody could have known at the moment of quoting — which is why a contingency is a planned item, not an emergency one.
Keep it as a separate budget line rather than a cushion in your head. A written line forces a decision every time you reach for it — "is the English version worth fifteen per cent of the budget" is a different question from "could we also add". The first sometimes gets a no, and that is healthy; the second almost never does.
If the contingency goes unused, it was not unnecessary. It means the scope was well described — that the work you put into the enquiry came back.
A good quote is short and specific, not long and general. Seven items tell you it is a document rather than a figure in an email:
A quote with no exclusions that does not say who supplies the content is not cheaper than one that does. It is less finished, and the difference is paid later.
The least-checked part of the document, and the one that decides what happens when the engagement ends. Four things worth stating explicitly before you negotiate price.
Whose name the domain is registered in. The only one of the four that can be impossible to recover without the other party's consent. The company, not the supplier.
Whose name the hosting account is in, and who gets the renewal notices. The commonest cause of a lost domain is not financial — it is an email to the mailbox of somebody who no longer works there.
Whose account the theme and plugin licences sit on, and what happens when the contract ends. The site keeps working but stops receiving updates — including security updates.
Who has administrator access. Not editor — administrator. If only the supplier does, every future change of supplier starts with a negotiation about access.
Separately, agree what happens to the design files — wireframes and graphic sources. They are sometimes treated as the supplier's working tools rather than part of the delivery, and that is to be settled rather than assumed either way.
The word "warranty" appears in most quotes and almost never means the same to two suppliers. Three things worth settling.
What is a defect and what is a new feature. A form that does not send is a defect. A form that sends but needs an extra field is a change of scope. The boundary is obvious only once written down.
For how long. The period is often counted from acceptance and is often shorter than the time it takes defects to surface — some only appear at the first real traffic or the first update.
Whether it covers the effects of updates. The most important question and the most often skipped: if an automatic plugin update breaks the layout six months after handover, is that a warranty defect or paid work? Both answers are honest — only the absence of one is not.
Two things to write into the contract before you start negotiating the amount. A contingency of 15–20% of the budget as a separate line — and the rules for billing changes of scope. Migrations and non-standard integrations cost more than assumed more often than they cost less, and the moment to agree the rate for them is before signing, not during.
A quote describes the build. It ends on publication day — the bill does not.
Hosting and the domain — whose choice has cost consequences of its own — licence renewals, updates and work on whatever breaks start accruing the day after launch and appear in no project quote, because formally they are not part of one. That is not concealment; it is a different document.
The consequence when comparing offers: two quotes can be identical and the three-year bill can still differ by a factor of two, depending on what the site stands on and how many licences it needs. The four layers of that bill, with renewal prices, we set out separately. Worth asking about at the quoting stage even though it is not part of it — an answer of "it depends" is information in itself.
Negotiating a quote is associated with one thing — pushing the rate down. That is where the room is smallest and the cost of a concession highest. Below is what can actually be moved.
The hourly rate — least of all. The supplier calculated it from the cost of the team and rarely moves more than a few per cent. A concession here comes back as a less experienced person on the project or a slower response — not because anyone is cheating, but because cheaper is done differently, not faster. Worth knowing before you push: rates in the two largest continental markets are falling, not rising — French JavaScript day rates down 10.6% to 16.6% year on year, and the German freelance average declining for the first time since its survey began.
Scope — most of all. The right place for the budget conversation. Cutting a third integration, two page templates or the photo shoot from stage one moves the figure genuinely and reversibly — each can be added later to a finished foundation.
The payment schedule — almost always. More instalments, the last one behind acceptance, instalments tied to milestones instead of dates. For the supplier that is cash flow, not margin, so the conversation is easier than it looks.
Warranty and a post-launch fixes package — worth it. One month or three for bug fixes, a pool of hours for small changes in the first quarter. Cheaper agreed before signing than bought afterwards in emergency mode.
A response time written as a number — worth it, if the site earns. "Support included" with no time attached is not a commitment. A specific response time for a critical outage, in the contract, is — sometimes the only difference between an hour and a week of downtime.
Transfer of rights and access to the code — not negotiable. The one item you have to have in full or not at all. Without a clause transferring the economic copyright, and without access to the code and the database, you cannot commission changes from anyone else and cannot move the site. With a supplier who disputes that, you are not negotiating price — you are negotiating whether you buy at all.
The order: rights and access first, then warranty and response time, then scope. Leave the rate alone — if the budget does not close after those three, the problem is scope, not the hourly price.
Regardless of the amount, several things should prompt a question. All concern the document, not the price.
One line for the whole thing. "Website build — €8,000" with no breakdown means every conversation about a change starts by establishing the scope from nothing, because it never was.
No exclusions. A document saying only what is included shifts the whole burden of interpretation to the moment of a dispute.
A deadline with no reference point. "Six weeks" from what? If from signing, and materials arrive in week three, the delay is formally yours.
No deposit. It sounds favourable and can signal that the supplier does not plan to start with the stage requiring their work before your approval — or that the project is filler.
Payment in full up front. The opposite extreme and more dangerous — after the transfer you have no leverage at all.
Silence about access. A quote that does not say whose name the domain goes in and who holds the administrator account defers that conversation to its hardest moment.
None of these means a bad supplier on its own. Each means a question worth asking before signing, not after.
A low figure is not itself a warning — often it is simply a narrower scope, and for some companies the right choice. The warning is a low figure against a broad description.
If the document promises a design from scratch, integrations, content and care, and the figure sits at the floor of the landing-page band on published price lists, one of those things is not what it appears to be. Usually "design" means choosing colours in a ready-made theme, and "content" means pasting across whatever is on the old site.
Three questions settle it faster than any analysis: how many people work on the project, how many of those stages have you seen this supplier deliver, and what happens when the scope turns out to be larger. The third matters most, because with an underpriced quote that moment always arrives.
And one thing worth saying plainly, because it cuts both ways: a quote well above the market also needs an explanation — just an easier one to get, because a supplier charging more usually knows why and will name it.
Four questions, asked before you look at the amounts.
A separate decision worth taking in parallel is whether you pay for the build once or spread it into a subscription — we costed both over five years. If the difference still looks unjustified after those four questions, it probably sits outside the document — in who is quoting, what they build on and where they work. That is a separate subject.
Safer for the budget, yes — you know the figure up front. Cheaper, not necessarily. A supplier quoting a fixed price takes on the risk of unknown requirements and prices it in. With a well-described scope, hourly or staged billing often comes out cheaper, because that margin disappears.
Because without a scope every number is a guess, and a guess that comes in too low becomes a dispute later. The reverse holds too: the more detail in the enquiry, the lower the quote tends to be — not out of courtesy, but because the margin for unknowns disappears.
The ratio — 30/40/30 or anything else — matters little. What matters is the condition attached to each instalment. One with no acceptance condition is a payment date rather than a milestone, and gives you no control over what you accept.
Yes, but scope negotiates better than the rate. Cutting the price without changing the scope means the same work for less — usually faster and less carefully. Ask instead what can be removed or deferred to a second stage: that saves money without damaging what remains.
In practice 15–20% of the build budget, and it goes on three things: waiting when content is not ready, scope added mid-project, and licences and integrations nobody planned for. None is a pricing error — they are things nobody could have known at the time.
One page: a list of pages by name, three sites you like, who writes the copy and takes the photographs, the names of the systems to integrate with, and the deadline with its reason if fixed. Without that, every supplier prices their own imagination and whoever imagined least wins.
Usually not, and that is not concealment — it is a different document. Hosting, domain, licence renewals and work on defects start accruing the day after launch. Two identical quotes can produce three-year bills differing by a factor of two, depending on what the site stands on.
We will say what is missing from it, where the risk margin sits and what to ask before you sign. Thirty minutes — including when the quote is from somebody else.
Build and upkeep are two separate bills. Market medians, our own starting rates, and eight articles — one for each question people ask about cost.
The same brochure site gets quoted at both ends of the range, and both prices can be honest. Six factors that decide which end you are quoted at.
The lowest quote is not the price of a website, only the smallest part of the bill. Three price tiers, the real cost after a year, four warning signs.
A free site is a real option with a precise limit. Three routes, what each one gives you, what it withholds, and what it costs once a year has passed.
The upkeep bill has four layers of differing predictability. Two you can price before you sign, two you cannot. With renewal prices from published lists.
A subscription is cheaper on day one and dearer from month 26. Where the two models cross, the five-year total for each, and what you own at the end.
What a free plan actually gives you, and what happens when the provider switches off. With 000webhost, Freenom and Weebly as the worked examples.
Registration is the advert, renewal is the price. What a domain and business hosting really cost across twelve months, with a source for every figure.
Your Partner in Business, Digital Vantage Team
Digital Vantage team is a group of experienced professionals combining expertise in web development, software engineering, DevOps, UX/UI design and digital marketing. Together we carry out projects from concept to implementation - websites, e-commerce stores, dedicated applications and digital strategies. Our team combines years of experience from technology corporations with the flexibility and immediacy of working in a smaller, close-knit structure. We work in agile methodologies, focus on transparent communication and treat each project as if it were our own business. The strength of the team is the diversity of perspectives - from systems architecture and infrastructure, frontend and design, to SEO and content marketing strategy. As a result, the client receives a cohesive solution where technology, aesthetics and business goals go hand in hand.
Rate this article
Back to the guide: Websites - a guide for entrepreneurs

Three layers in the order that matters, the list of checks, and the price stated outright. With three findings an owner will never spot on their own.

Indexing and ranking run on two different clocks. The four gates a site passes through, with the times measured on our own corpus rather than quoted.

The same brochure site gets quoted at both ends of the range, and both prices can be honest. Six factors that decide which end you are quoted at.

The lowest quote is not the price of a website, only the smallest part of the bill. Three price tiers, the real cost after a year, four warning signs.

A free site is a real option with a precise limit. Three routes, what each one gives you, what it withholds, and what it costs once a year has passed.

Build and upkeep are two separate bills. Market medians, our own starting rates, and eight articles — one for each question people ask about cost.

Four pricing mechanisms hidden in builder plans, what the second year actually costs, and what you can export when you outgrow the tool.

Four channels, when each one works, and what is left when you stop paying. With twelve months of our own traffic split, and the cost per enquiry.

Learn how to double your conversions without additional budget: 2% benchmark, 3s rule, 5 free tools, mobile 60% conversion and 90 day plan.