What a Small Business Website Really Costs.

The six factors that actually move the price

Almost every difference between a cheap website and an expensive one comes down to how much bespoke human work is involved. Templates are cheap because the design decisions have already been made and paid for once. Custom work is expensive because it is made for one business only.

When you compare quotes, these are the variables that explain the gap. If a quote is unusually low, one of these has almost certainly been reduced or removed.

• Design origin. A pre-built theme with your logo dropped in sits at one end. A layout designed around your specific offer and audience sits at the other.
• Number of unique layouts. Ten pages that reuse three layouts is a fraction of the work of ten pages that each need their own structure.
• Who writes the copy. Writing persuasive copy for an unfamiliar business is slow, skilled work. Many quotes assume you supply finished text, and that assumption is often buried.
• Existing brand assets. A usable logo, real photography and a defined colour palette remove entire workstreams. Starting from nothing adds them back.
• Integrations. A contact form is trivial. Online booking, payments, inventory, a CRM handshake or a customer login are each separate pieces of engineering.
• Post-launch support. Hosting, updates, security patching, backups and content changes either cost money or cost your time. A quote that ignores them has moved that cost rather than removed it.

The four ways small businesses usually buy a website

Each of these routes has a legitimate use. The mistake is choosing one without understanding what it optimises for.

Do-it-yourself builders are the cheapest in cash and the most expensive in time. You are not paying for design labour, so you are supplying it. The output can be perfectly good for a simple business, but the hours are real and they come out of running your company.

Individual freelancers are flexible and often excellent. The risk is concentration: one person is simultaneously the designer, developer, project manager and support desk, and holidays, illness or a bigger client can stall your project.

Traditional agencies quote each project individually, usually after a discovery process. That is appropriate for genuinely complex requirements, and it is why costs are high and timelines are measured in months. Much of the price is the bespoke process itself, not the finished pages.

Productised services, which is the model Nine99 uses, fix the scope and the price in advance. The efficiency comes from doing a well-defined thing repeatedly rather than from cutting corners. The trade-off is honest: you get a defined scope rather than an open-ended one.

What is usually excluded from a website quote

Exclusions cause more disputes than the headline price ever does. Before you sign anything, establish which of these are inside the number and which arrive later as separate invoices.

• Domain registration and annual renewal.
• Hosting after any introductory period, and what the standard rate becomes.
• An SSL certificate, though this is now bundled by most credible hosts.
• Copywriting, if you assumed it was included.
• Stock or original photography, and any licensing.
• Revisions beyond a stated number of rounds.
• Search optimisation beyond basic on-page fundamentals.
• Ongoing content updates once the site is live.
• Third-party software subscriptions such as booking or email tools.

The cost that continues after launch

A website is not a one-off purchase in the way a printed brochure is. Software dependencies age, browsers change, and content goes stale. Budgeting only for the build is the most common planning error we see.

Hosting is the obvious recurring item. Less obvious are security updates, dependency upgrades, backups and periodic content changes. If nobody is doing that work, it is not free, it is simply deferred until something breaks.

There is also a performance dimension with real consequences. Google publishes Core Web Vitals as a public set of user-experience metrics covering loading, interactivity and visual stability, and treats page experience as a ranking consideration alongside content relevance. A site that degrades over time can quietly lose visibility, which is a cost even though no invoice is issued.

Accessibility follows the same pattern. The Web Content Accessibility Guidelines are published by the W3C as a formal standard, and building to them from the start is far cheaper than retrofitting a finished site.

How to compare two quotes fairly

Because scope varies so much, comparing headline prices tells you very little. These questions surface the differences quickly, and any competent provider will answer them without hesitation.

• How many unique page layouts are included, as opposed to total pages?
• Who writes the copy, and what happens if I do not supply it?
• How many revision rounds are included, and what does an extra round cost?
• What is the hosting rate once any introductory period ends?
• Who owns the finished site and its content?
• What is the timeline, and what specifically starts the clock?
• What happens if I need a change six months after launch?

What you are actually paying for

It helps to think in hours rather than deliverables, because hours are what a provider is really selling. A website that looks simple from the outside can involve a surprising amount of separate work, and understanding where the time goes makes a quote far easier to interpret.

Discovery comes first: understanding what the business sells, who buys it and what a visitor needs to see before making contact. Skipping this is the single most common reason a website looks acceptable but generates nothing. A site built without it tends to describe the company rather than address the customer.

Then design, which is where the largest cost variation sits. Adapting a template is measured in hours. Designing original layouts around a specific offer is measured in days, because every decision about hierarchy, emphasis and flow has to be made from scratch rather than inherited.

Then build, where the visible design becomes working code that behaves correctly on a phone, a tablet, a laptop and a wide monitor, in several browsers, at different text sizes, for keyboard and screen reader users. This is the stage most often underestimated by buyers, because none of the work is visible when it has been done properly.

Then content assembly, testing and launch: writing or placing copy, preparing images at sensible file sizes, configuring forms so submissions actually arrive, connecting analytics, checking links and setting up hosting and certificates. Individually small, collectively a substantial block of time.

Where cheap websites usually go wrong

A low price is not automatically a problem. A low price that hides one of the following usually is, and each of these is recoverable only by paying somebody to redo the work.

• No forms tracking. The site generates enquiries but nobody knows which pages produced them, so there is no way to tell what is working.
• Images uploaded at full camera resolution. The pages load slowly on a phone, which is where most visitors are, and the cause is invisible to anyone not looking for it.
• A template so heavily loaded with unused features that performance suffers permanently.
• Copy written to describe the business rather than to answer what a customer wants to know before enquiring.
• No analytics installed at all, which makes every future marketing decision guesswork.
• A proprietary platform you cannot export from, so improving the site later means starting again.
• No mobile testing beyond resizing a desktop browser window, which misses touch targets, real device performance and how forms behave on a small keyboard.

Where the Nine99 approach sits

Nine99 charges a fixed 999 dollars for its website package, with three months of hosting included. The price is fixed because the process is standardised, not because the work is minimal or offshored.

The build window is three business days, and it begins when onboarding information, content and any required account access have been received rather than at the moment of purchase. That distinction matters: the constraint on most website projects is not development speed, it is waiting for material from the client.

This model deliberately trades open-ended scope for speed and price certainty. If your requirements genuinely include a complex integration, a bespoke application or a multi-location content structure, a fixed-scope package is the wrong tool and an honest provider will say so rather than sell it to you anyway.

One last point worth making regardless of who you hire. Ask any provider what happens after launch, and get the answer in writing before you pay. Who hosts the site, what the ongoing rate becomes, who applies security updates, how you request a change, and how quickly it gets done. A website is a running system rather than a finished object, and the cheapest quote is rarely the cheapest arrangement once a year of small changes has accumulated.

The three performance numbers that decide whether a site works

Cost conversations tend to focus on how a website looks, because appearance is the part a buyer can judge instantly. Performance is harder to see and matters more, because it affects both what visitors do and how search engines treat the page.

Google publishes three Core Web Vitals with specific thresholds, and they are public rather than a matter of opinion. Largest Contentful Paint, which measures when the main content becomes visible, should occur within 2.5 seconds. Interaction to Next Paint, which measures responsiveness to a tap or click, should be 200 milliseconds or faster. Cumulative Layout Shift, which measures how much the page jumps around while loading, should be 0.1 or lower.

These numbers are useful in a pricing conversation because they are testable. Anyone can run a competitor site or a prospective supplier’s own portfolio through Lighthouse or PageSpeed Insights and see the result. A provider whose own website fails its thresholds is telling you something about the sites it builds.

The reason this belongs in a cost discussion is that performance is cheap to build in and expensive to add later. Image handling, font loading, script deferral and layout stability are decisions made during construction. Retrofitting them means revisiting work that has already been paid for once.

• Ask for a Lighthouse or PageSpeed score on a site the provider has actually built, not a demo.
• Ask on mobile rather than desktop. Mobile is where most small business traffic arrives and where slow sites fail.
• Treat a refusal to show a score as an answer in itself.

Accessibility is a cost factor, and retrofitting is the expensive route

The Web Content Accessibility Guidelines are published by the W3C as a formal standard, currently at version 2.2. They cover things like colour contrast, keyboard operability, meaningful alternative text for images, visible focus indicators and correctly labelled form fields.

Accessibility rarely appears on a website quote, which is precisely why it is worth raising. Built in from the start it is close to free, because it mostly consists of doing standard things correctly: using real headings in order, using real buttons rather than styled div elements, choosing colours with sufficient contrast, and labelling inputs. Added afterwards it becomes a remediation project, because the underlying markup has to change.

There is a commercial dimension beyond compliance. The same practices that make a site usable with a keyboard or a screen reader also make it more robust for everyone: clear structure, predictable navigation, forms that explain their errors. Sites built this way tend to convert better, not because accessibility is a growth tactic, but because clarity is.

If a quote does not mention accessibility at all, ask which standard the build targets and how it will be checked. A provider who has thought about it will answer in a sentence. A provider who has not will change the subject.

The real number is three-year cost, not the launch invoice

Comparing website quotes on the launch price alone is the most common mistake in this purchase, because it compares the smallest part of the cost. A website is a running system with recurring obligations, and those obligations continue whether or not anyone is managing them.

A more honest comparison is to add up everything you will pay over three years across every route you are considering. That means the build, plus hosting, plus the domain, plus any platform or plugin licences, plus support or maintenance, plus the cost of the changes you will inevitably want, plus the value of your own time if you are the one making them.

This exercise frequently reverses the ranking. A do-it-yourself builder with a low monthly fee can total more than a fixed-price build once three years of subscription and your own hours are counted. An expensive agency build with everything included can total less than a cheap build that generates a support invoice every month. The launch price simply does not predict the total.

The point is not that one route wins. It is that the number people compare is not the number they pay. Doing this arithmetic before committing takes about twenty minutes and is the single highest-value thing a buyer can do.

• Build or setup cost, once.
• Hosting, monthly or annually, at the standard rate rather than any introductory rate.
• Domain registration and renewal.
• Platform subscription, theme licences and any paid plugins or apps.
• Security updates, backups and patching, whether paid for or done by you.
• Content changes, at whatever the provider charges or whatever your time is worth.
• The redesign you will want in three to five years, because designs date.

Platform choice is a cost decision disguised as a technical one

Buyers are often told the platform is a detail to leave to the professionals. In cost terms it is one of the most consequential decisions in the project, because it determines what future changes cost and whether you can leave.

Open platforms such as WordPress, or framework-based builds using something like Next.js, produce a site whose files and content you can move to another provider. Hosted proprietary platforms are frequently easier to launch and genuinely good at what they do, but the site exists inside that platform and cannot be exported as a working website. Some providers also build on their own private systems, which means the site only continues to exist while you remain a customer.

None of these is dishonest by itself. What matters is knowing which one you are buying, because the cost of leaving is paid at the worst possible moment: when you are already unhappy. A site you cannot move is a site whose future price you do not control.

The practical test is a single question. If you stopped working with this provider tomorrow, what exactly could you take with you, and would it still function? Ask it before you pay, and get the answer in writing rather than in a phone call.

What no website quote can honestly promise

A meaningful part of what makes website pricing confusing is that some of what gets sold cannot be delivered by anyone at any price. Recognising those claims is a practical way to sort credible quotes from optimistic ones.

Nobody can guarantee a search ranking. Google’s published guidance is explicit that no one can guarantee placement, and search results are produced by ranking systems no provider controls. Nobody can guarantee a specific volume of leads from a website alone, because that depends on demand, competition and pricing in your market. Nobody can guarantee a conversion rate, because conversion is a property of your offer as much as your web pages.

What a competent provider can commit to is process and specifics: what will be built, to what standard, by when, what happens if it slips, who owns the result, and what ongoing costs apply. Those are all verifiable. Outcomes are not, and a quote that leans on guaranteed outcomes is usually compensating for vagueness elsewhere.

How to run a fair comparison in one afternoon

You do not need technical knowledge to buy a website well. You need the same information from every supplier, which is harder than it sounds because quotes are rarely structured the same way.

Send every provider an identical short brief: what your business does, roughly how many pages you expect, whether you have copy and photography, what has to work such as a booking flow or payments, and when you need it live. Ask each to reply against the same fixed list of items so the responses can be laid side by side.

Then read for exclusions rather than inclusions. Any quote can list impressive features. The distinguishing information is what has been left out, what is described in vague language, and what carries a separate price later. Where two quotes differ sharply on one line, ask both to explain that line specifically.

Finally, weight the answers about after launch as heavily as the answers about the build. Most of the money and nearly all of the frustration in website ownership happens after the site goes live, and it is the part almost nobody asks about while comparing prices.

• Total price, and precisely what is inside it.
• What you must supply, and what happens to the timeline if you are late.
• Who writes the copy and who sources the images.
• Number of revision rounds included, and the cost of further ones.
• Hosting rate after any introductory period.
• Who owns the design, the content, the domain and the code.
• How you request a change after launch, and the response time.
• What you can take with you if you leave.

Payment terms are part of the price

Two quotes with an identical total can carry very different risk depending on when the money changes hands. Staged payments tied to deliverables protect both parties: the supplier is not funding the work indefinitely, and the buyer is not fully paid up before anything exists.

Ask what triggers each payment, what happens if the project stalls because material has not arrived from you, and what happens if you cancel midway. A deposit that becomes non-refundable the moment it is paid is a materially different product from one that covers work already completed, even where the headline figures match exactly.

Fixed-price productised services usually take payment up front, which is reasonable when the scope is genuinely fixed and the timeline is short. The safeguard in that model is not a payment schedule but a clearly written scope document, so read that document carefully rather than negotiating the invoice.

What most sources get wrong

Each claim below is one we found published and being repeated. The correction is drawn from primary documentation.

The Nine99 view

Our own judgement, not drawn from the sources above. Treat it as opinion formed in practice rather than documented fact.

The most useful predictor of a successful website purchase is not the budget. It is whether the buyer has established, in writing and before paying, who supplies the copy. Copy is the workstream most often assumed away by both sides: the provider prices on the assumption text will arrive, the client assumes writing is included. Projects that stall almost always stall there rather than in development.

Exclusions deserve more scrutiny than inclusions when reading a quote. Any supplier can list features, and feature lists are close to free to produce. What differentiates quotes is what has been left out, what is described in deliberately soft language, and what will arrive later as a separate invoice. Reading a quote for its silences is faster and more diagnostic than comparing its bullet points.

A fixed-price model should be judged on whether the scope genuinely fits, not on the price. Nine99 fixes both because standardising production removes the cost of quoting each job individually. That is a real efficiency and it is also a real constraint: a business needing a bespoke integration, a custom application or a large multi-location content structure is better served elsewhere, and should be told so rather than sold a package that does not fit.

Common questions