How Much Does a Website Cost in 2026?

Quick answer

A business website can cost anywhere from a few hundred dollars to well over $150,000. The useful question is not where that enormous range starts and ends, but which part of it your project actually belongs in.

As a practical 2026 planning guide, a self-built site on a platform such as Wix, Squarespace, or Webflow can have little or no professional build cost beyond the platform, domain, assets, and your own time. A professionally configured template site commonly falls around $1,200 to $5,000. A custom small-business website sits roughly in the $3,000 to $18,000 range. A conversion-focused marketing or content site with custom design and integrations often lands between $10,000 and $50,000. Professionally delivered ecommerce projects can start around $5,000 and move past $70,000 once catalog complexity, inventory, tax, subscriptions, or ERP integration enter the picture. A focused first version of a custom web application can start around $18,000, while complex products can comfortably exceed $150,000.

Those are planning bands, not universal market averages or quotations. Geography, scope, seniority, content, delivery model, and the systems being integrated can move a real quote in either direction. The rest of this guide explains why two legitimate quotes for what sounds like “the same website” can differ by 5x or 10x, and how to tell whether that difference represents real scope or just a poorly defined proposal.

Website cost at a glance

Every row below is an interval, not a hard boundary. A lean scope or a lower-overhead independent developer can land below the floor shown. A complex scope, difficult integration, accelerated timeline, or larger delivery team can move beyond the ceiling.

The recurring-cost column should be read as a typical recurring budget for that delivery model, not a universal subscription price. Depending on the category, it can include platform fees, hosting, maintenance, support, or infrastructure. It does not assume ongoing advertising or content production, and payment-processing fees are additional for businesses that sell online.

Delivery model	Realistic initial cost	Typical recurring cost	Typical scope	Best fit	Important limitation
DIY website builder (Wix, Squarespace, Webflow, self-built)	$0–$1,500	$20–$70/month	5–15 template pages, self-edited	Simple, single-location businesses; pre-revenue validation	Platform lock-in, limited code portability, and a ceiling on some custom functionality
Professionally configured template site	$1,200–$5,000	$50–$200/month	5–12 pages on an existing theme, professionally set up	Small businesses that want a credible site fast without bespoke design	Still visually and functionally template-bound underneath
Custom small-business website	$3,000–$18,000	$150–$500/month	8–20 pages across several unique templates, content architecture, custom visual design	Businesses that depend on being found and trusted online	The design can outrun the content budget if you're not careful
Custom marketing/content site	$10,000–$50,000	$300–$1,200/month	15–40 pages, many unique templates, CRM/marketing integrations, blog or resource center	Growth-focused B2B or professional-services businesses	Cost tracks unique templates and integrations, not page count
Ecommerce site	$5,000–$70,000+	$100–$3,000+/month, plus payment processing	Product catalog, checkout, tax/shipping, accounts	Businesses selling online	Platform and app costs recur; payment-processing fees scale with transaction volume
Custom web application	Highly variable, typically $18,000–$150,000+ for a focused first version	Planning heuristic: ~15–25% of build cost/year, plus infrastructure	Custom data model, workflows, auth, integrations	Products where the site is the business, not a brochure for it	Page/template thinking doesn't apply; scope is the whole story

How I arrived at these ranges

There is no authoritative database that can tell you what “a website” costs, because the word covers everything from a five-page brochure site to a multi-tenant software product. I built these bands by triangulating three kinds of 2026 data: official platform pricing, current freelance and agency rate data, and the amount of design, development, content, integration, and project work a realistic scope requires.

The lower ends of the professional ranges assume a lean delivery model, often an experienced independent developer or a small specialist team. The higher ends increasingly overlap with agency delivery, more disciplines working in parallel, greater project-management overhead, deeper design work, and more complex requirements. Current Clutch web-design data shows just how wide the agency market is by location, while its separate web-development pricing data shows why one global hourly rate would be misleading. Upwork’s current web-developer pricing guide shows the same spread on the freelance side.

The custom web application row is deliberately wide. A booking tool for one location and a multi-tenant SaaS platform both qualify as custom web applications, but they are not remotely equivalent scopes. Once custom data models and workflow logic become the product, a page-count pricing model stops being useful.

Why “how much does a website cost” is the wrong first question

“Website” describes so many different products that comparing their prices on that word alone is close to meaningless. A five-page site for a local plumber, a lead-generation site for a professional-services firm, a full ecommerce operation, and a SaaS product are all technically websites. They have almost nothing else in common.

The plumber’s site needs to explain what the business does, where it works, and how to get in touch, while looking credible enough that a stranger trusts the company with a job inside their home. A lead-generation site has a different job: turn a cold visitor into a qualified conversation. That can require content strategy, case studies, trust signals, conversion paths, and a CRM connection that the brochure site simply does not need.

An ecommerce operation adds product data, variants, tax, shipping, inventory, customer accounts, payment processing, and checkout behavior. A web application can add authentication, permissions, databases, business rules, integrations, background jobs, and interfaces that have to remain correct while the product evolves for years.

Asking what a website should cost without specifying which of those you are building is like asking what a vehicle should cost without saying whether you mean a bicycle, a delivery van, or a freight truck. The category name is the same. The underlying product is not.

What actually determines website cost

Once you know roughly which category you are in, the price inside that category is driven by a fairly short list of variables. These are the ones worth understanding before you compare quotes.

Unique templates matter more than page count

This is one of the most useful distinctions in website pricing.

A site with 50 pages built from three reusable templates, for example a standard content page, a service page, and a location page, can be relatively efficient to design and develop. The expensive work is creating the reusable system. Once that exists, additional pages built from it are increasingly a content-production task rather than a new engineering task.

A site with 15 pages that require ten different layouts is a different project. A homepage, pricing comparison, case-study layout, interactive calculator, team directory, resource hub, and several campaign-specific pages all require their own design decisions, responsive behavior, CMS modeling, and testing.

If you take one practical question from this article, make it this one: how many unique templates does the quote assume?

Two proposals for a “20-page website” can differ dramatically because one assumes four reusable templates and the other assumes twelve. The headline page count does not expose that difference.

Custom design and branding

There is a real spectrum here, and custom design is not automatically the right choice at every point on it.

A prebuilt theme used close to its default structure, with typography, colors, imagery, and branding adjusted, is the cheapest route and can be entirely appropriate. A customized theme goes further by changing layouts, spacing, components, and interaction patterns while retaining the underlying theme. A component-based custom design system is built around the brand and the actual content model. Bespoke brand and UX work goes further again by defining or reworking the visual identity and experience as part of the project.

The expensive option is not inherently the good option. A well-chosen theme executed properly can outperform a badly designed bespoke site. Pay for custom design when differentiation, conversion, usability, or long-term flexibility creates enough value to justify it.

Content is often a bigger project than expected

Copywriting, photography, video, diagrams, product data, case studies, and migration from an old site are routinely underestimated because they do not look like “development” on a project plan.

Formatting and SEO editing add another layer. Raw copy in a document is not the same deliverable as content that has been structured into useful headings, summaries, calls to action, metadata, internal links, and reusable CMS fields.

If a proposal says “content supplied by the client,” treat that as a real project dependency. Ask whether the content exists, who is responsible for producing it, who approves it, and when it will be ready. Development can be complete while a site still cannot launch because photography, product data, legal copy, translations, or final approvals are missing.

If you are deciding how much content is actually worth producing in a search environment increasingly shaped by AI answers, how content websites can survive AI Overviews covers the distinction between commodity content and material that still gives people a reason to visit the source.

The CMS is a workflow, not just software

“We’ll use WordPress” tells you very little about implementation quality or project cost. The same applies to most content-management systems.

At one end, a developer can configure an existing theme and give editors a largely freeform page builder. At the other, the site can have a deliberate content model where services, locations, case studies, people, products, and related content are structured as reusable data with controlled editing options.

The second approach takes more thought up front, but it can make the website cheaper to operate because staff can publish and update content without rebuilding layouts or calling a developer for routine changes.

A CMS that is painful to use becomes an operational cost after launch. Editors either need developer help for simple changes or start working around the system until the design becomes inconsistent. That is why CMS work belongs in the scope discussion rather than being reduced to the name of the platform.

If you are weighing WordPress specifically, I have also written about why WordPress remains a strong CMS choice when the implementation is structured deliberately rather than assembled as an uncontrolled collection of plugins and page-builder blocks.

Development complexity

This is the category most people picture when they think about website cost, but even here the range is wide.

A marketing site still needs responsive behavior, accessible components, forms, analytics, performance work, browser testing, CMS integration, deployment, and a maintainable frontend. Move beyond that and the project can add authentication, databases, APIs, search, dashboards, user roles, conditional forms, background processing, payments, or custom workflows.

That is usually the point where the conversation shifts from “website” to “web application.” Once the product contains business logic, the complexity of that logic matters more than the number of screens.

Third-party integrations

CRM platforms, booking engines, marketing tools, ERPs, property-management systems, payment providers, analytics products, inventory tools, and internal APIs are common dependencies on business websites.

The important pricing detail is that integration complexity depends heavily on the system being integrated, not simply the number of integrations on a list.

A narrow integration against a well-documented REST API with a sandbox, good authentication support, clear webhooks, and a maintained SDK can be straightforward. The same business requirement against an undocumented legacy system, inconsistent data, awkward authentication, or an API with no useful test environment can become a substantial piece of work.

That is why “CRM integration included” is not enough detail in a quote. The proposal should identify what data moves, in which direction, how errors are handled, whether synchronization is real-time or scheduled, and what the third-party API actually supports.

Ecommerce

Ecommerce cost is driven by much more than the presence of a checkout button.

Catalog size, variants, subscriptions, tax, shipping, currencies, inventory, ERP integration, customer accounts, search, merchandising, discounts, returns, and how far checkout needs to deviate from platform defaults can all move the budget.

A 40-product store selling in one country and one currency is a different engineering problem from a 4,000-SKU catalog with real-time inventory synchronization, wholesale pricing, localization, and tax rules across several markets. Both are ecommerce websites, but they should not be priced as the same product.

Accessibility

Accessibility is part of engineering quality, not a widget that can be added at the end. Semantic HTML, keyboard navigation, focus management, form labeling, contrast, error handling, target sizes, and accessible authentication patterns are decisions made throughout the build.

The current W3C recommendation is WCAG 2.2, and W3C encourages organizations to use the latest version. Formal legal obligations can still refer to earlier versions or other standards, so the engineering target and the legal requirement are not always worded identically.

For businesses operating in the EU, the European Accessibility Act has applied to defined categories of covered products and consumer services since 28 June 2025. The scope includes areas such as ecommerce, consumer banking, certain passenger-transport services, electronic communications, and e-books. It is not a blanket rule that every private website in Europe has the same obligations, and the directive also contains exclusions, including an exemption for microenterprises providing services. National implementation and the specifics of the business matter.

For the technical side, W3C notes that the European EN 301 549 standard currently uses WCAG 2.1 while the next version is expected to move to WCAG 2.2. If legal compliance is material to your business, confirm the applicable requirement in your jurisdiction rather than treating a blog article as legal advice.

From a project-cost perspective, the useful point is simpler: accessibility is usually cheaper to design and test as part of the normal build than to retrofit after inaccessible patterns have been repeated across every template. How much it changes the budget depends on the starting point, complexity, testing requirements, and compliance obligations.

Multilingual websites

A genuinely multilingual site is an architecture decision, not just a translation job.

It affects URL strategy, content modeling, metadata, hreflang, navigation, search, localized assets, editorial workflow, and what happens when the source language changes after translations have already been approved.

Machine translation can be useful inside that workflow, but swapping the words is not the same thing as building the underlying multilingual system. A business expecting meaningful organic traffic or sales from several markets needs to decide how content is created, reviewed, versioned, and published in each language.

Performance and technical SEO

There is a real difference between installing an SEO plugin and building a site that is fast, crawlable, internally coherent, and technically easy for search engines to understand.

A plugin can help generate sitemaps or expose metadata fields. It cannot by itself fix an oversized JavaScript bundle, duplicate URL architecture, slow server responses, poor image handling, broken canonicals, inconsistent internal linking, or templates that generate low-value pages at scale.

Performance and technical SEO are therefore part of implementation quality, not magic switches at the end of the project. If you want to see how I approach performance beyond one homepage Lighthouse score, Unlighthouse is useful because it audits patterns across an entire site rather than one URL at a time.

Cost by delivery model

The same scope can be delivered in very different ways. Who does the work affects price, communication, capacity, and risk.

DIY website builders

Wix, Squarespace, and Webflow are capable products, and for the right project they are the rational financial choice.

As of August 2026, Wix lists annual-billed US reference prices around $17/month for Light, $29 for Core, and $39 for Business, with higher tiers available. Webflow currently lists Basic at $15/month and its new Premium Site plan at $25/month when billed yearly. Squarespace’s current Basic, Core, Plus, and Advanced lineup is available on annual or monthly billing, with current US annual pricing around $19, $29, $49, and $99 per month respectively as of August 2026. Check Squarespace’s live pricing page because prices and currencies vary by market.

This is the right route when you are validating an idea, launching a simple local business, building a temporary campaign site, or genuinely do not need custom behavior. It can remain the cheapest option for the full life of the site when requirements stay simple.

The trade-off is not that you “do not own your website.” The more accurate issue is platform dependence and portability. You can own your domain, brand, content, imagery, and other business assets while still being dependent on a proprietary platform for the underlying site implementation and certain functionality. Whether that matters depends entirely on what you expect the site to become.

For a deeper decision framework, see AI website builder vs web developer.

Template-based WordPress or similar CMS builds

“WordPress website” hides an enormous range.

A professionally configured site based on an established theme, sensible plugins, standard content types, and a small amount of customization can be delivered efficiently. A bespoke WordPress project with custom components, structured content types, custom fields, editorial constraints, integrations, performance work, and a tailored theme is custom development even though the CMS happens to be WordPress.

Both can honestly be sold as “a WordPress website.” They are not the same deliverable, so they should not be expected to cost the same amount.

Freelancers

A freelancer gives you direct access to the person doing the work, usually with fewer organizational layers and lower fixed overhead than a larger agency. For a well-scoped project that fits one person’s skills and capacity, that can be extremely efficient.

The market is also very wide. Upwork currently shows a $15 to $50/hour median range for web developers, while its own experience breakdown extends from roughly $15 to $25/hour for entry-level work to $50 to $200/hour for expert work. Those numbers are marketplace observations, not a quality scale. Geography, specialization, reputation, availability, communication, and the type of project all matter.

The main trade-offs are capacity and continuity. One person can only do so many things in parallel, and a project that simultaneously needs deep brand design, frontend development, backend architecture, copywriting, QA, and project management may be a poor fit for a single independent developer even if that developer is excellent at several of those disciplines.

For a smaller custom website, however, direct access to an experienced generalist can remove handoffs and keep the commercial structure lean. That is one reason a legitimate independent quote can sit well below a larger agency quote without implying lower-quality work.

Agencies

An agency can bring several disciplines together under one contract and run work in parallel. That can mean dedicated design, engineering, content, QA, project management, and account coverage rather than expecting one person to carry the entire project.

Current Clutch web-design pricing data shows how strongly geography affects agency rates: roughly $100 to $149/hour for firms in the United States, Canada, and Australia; $50 to $99/hour in the United Kingdom; $25 to $49/hour in countries including Ukraine, Poland, Spain, and Mexico; and lower typical bands in some other markets. Clutch’s separate web-development dataset reports a platform-wide typical range of $25 to $49/hour, which is a good reminder that “agency rate” depends on the discipline and dataset being discussed.

Some agency cost is organizational overhead. Some of it also buys real resilience: multiple specialists, cover when somebody is unavailable, dedicated QA, project management, and the ability to move several workstreams forward at the same time.

That premium makes the most sense when the project is large enough, risky enough, or time-sensitive enough that parallel capacity and organizational continuity are valuable in their own right.

In-house teams

Hiring in-house starts to make sense when the website or web application is a permanent, evolving capability rather than a project that occasionally needs updates.

The latest US Bureau of Labor Statistics Occupational Employment and Wage Statistics data, for May 2025, reports a mean annual wage of $98,770 for web developers and a median hourly wage of $44.54. See the BLS May 2025 occupational wage data. Salary is only part of the employer’s total cost once benefits, payroll taxes, equipment, recruitment, management, and non-billable time are included.

That expense is rational when a team ships continuously, product knowledge needs to remain inside the company, or the cost of waiting for an external supplier becomes higher than maintaining internal capacity. For a marketing site that receives a few meaningful changes each quarter, a full-time development hire is usually difficult to justify on website work alone.

Four realistic project budgets

These are illustrative scenarios designed to show how the cost drivers combine. They are not quotations for real projects.

Example 1: Local service business

Scope: Home, About, Services overview, five individual service pages built from one reusable template, Contact, a simple CMS, analytics, basic technical and on-page SEO setup, responsive design, and standard forms. No ecommerce, booking engine, or complex integration.

Likely cost category: Professionally configured template site to entry-level custom small-business website, roughly $2,000 to $9,000 depending on design and content scope.

What drives the cost: Whether the visual design is theme-based or custom, how much photography and copywriting are required, and how many genuinely unique layouts exist.

What could reduce the budget: A strong existing visual identity, client-supplied final copy, a suitable premium theme, and a small number of reusable templates.

What could increase it: Several locations with their own content requirements, custom photography or video, booking functionality, or additional layouts introduced during the project.

Example 2: Professional services company

Scope: Everything in example 1, plus custom design, case studies, conversion-focused page architecture, CRM integration, multilingual support for two markets, and migration from an existing website.

Likely cost category: Custom small-business website to custom marketing/content site, roughly $10,000 to $40,000.

What drives the cost: The number of unique templates, multilingual content architecture, migration quality, design depth, and exactly what the CRM integration is expected to do.

What could reduce the budget: Launching one language first while keeping the architecture multilingual-ready, simplifying the CRM integration, or reusing a smaller component set across more pages.

What could increase it: Additional languages, complex content migration, a heavily customized CRM data model, extensive copywriting, or a brand identity project happening at the same time as the website.

Example 3: Ecommerce business

Scope: A few hundred products with size and color variants, hosted checkout, tax for a limited number of markets, one currency, customer accounts, and an integration with an existing accounting or inventory tool.

Likely cost category: Ecommerce, roughly $10,000 to $35,000 for this example. Platform subscriptions, apps, support, and payment-processing fees continue after launch.

What drives the cost: Product data quality, variant complexity, inventory synchronization, integrations, merchandising, and how far checkout and account behavior need to move beyond platform defaults.

What could reduce the budget: A smaller catalog, standard Shopify behavior, clean product data, limited localization, and manual processes where real-time integration is not yet commercially necessary.

What could increase it: Multi-currency support, subscriptions, wholesale pricing, loyalty systems, ERP integration, advanced search, or several tax and fulfillment jurisdictions.

Example 4: Web application

Scope: A booking or scheduling product with real-time availability logic, user accounts with different permission levels, payment, an admin dashboard, and a database-backed domain model.

Likely cost category: Custom web application, highly variable. A focused first version might reasonably sit around $25,000 to $110,000, while a broader or more regulated product can go much higher.

What drives the cost: Business rules, data integrity, permissions, integrations, testing, security requirements, and the number of edge cases the first version is expected to handle.

What could reduce the budget: A deliberately narrow first release with fewer roles, simpler workflows, and fewer integrations, followed by iteration based on actual usage.

What could increase it: Multi-tenancy, complex scheduling rules, audit requirements, compliance constraints, sophisticated payments, or dependencies on external systems that do not expose the capabilities the product needs.

Page-count thinking is almost useless in this category. A ten-screen application with difficult workflow logic can require more engineering than a forty-screen application that is mostly conventional forms and content views.

Initial cost versus total cost of ownership

The purchase price is not the same thing as the cost of owning the website.

A realistic ownership model can include the initial implementation, hosting, domain renewal, platform subscriptions, premium themes or plugins, monitoring, backups, maintenance, security work, content updates, future features, developer support, migrations, ecommerce apps, and payment-processing fees.

Not every website incurs every cost. A static site with a simple deployment model can require very little routine maintenance. A WordPress installation with many dependencies has a different operational profile. A custom web application that is actively evolving is different again.

The point is to compare the whole operating model, not just the invoice that gets the site online.

A three-year website cost example

Here is a comparison that does not try to make custom development win.

Imagine a small professional-services business that needs a credible five-to-eight-page site, a contact form, basic analytics, and light content editing.

Path A, DIY builder. A suitable Wix or Squarespace plan can sit around the low tens of dollars per month, depending on the feature tier and billing cycle. Budget roughly $20 to $40/month plus a domain where it is not included. Over three years, that can remain around $800 to $1,600 before optional add-ons if the requirements stay simple.

Path B, professionally built custom site. Assume an initial build around $6,000, on the lean end of the custom small-business range, plus roughly $150/month for a bundle of hosting and light maintenance/support. Over three years, that is about $11,400.

The builder wins the cash-cost comparison by a wide margin in that scenario. That is exactly why I would not recommend custom development simply because it is more flexible.

The custom project earns its higher cost only if the business actually benefits from things the cheaper route does not provide: stronger differentiation, custom conversion work, integrations, better ownership and portability, a more appropriate content model, unusual functionality, or an architecture that avoids a costly constraint later.

Even then, a future migration does not automatically make the builder the expensive choice. A business can use a builder cheaply for several years, outgrow it, migrate later, and still have made the correct financial decision at the beginning. Total cost has to be evaluated against what the business needed at each stage, not against a rule that custom development always wins eventually.

Hidden and commonly excluded costs

Not every site incurs every item here, but a responsible proposal should tell you which ones apply rather than letting them appear as surprises after launch.

Domain registration and renewal: A conventional domain is often inexpensive, but pricing varies substantially by TLD, registrar, and whether the first year is promotional.
Hosting: Basic shared hosting can cost only a few dollars per month. Managed business hosting, higher-availability infrastructure, or application hosting can cost much more depending on traffic, support, backups, and compute requirements.
Business email: Google Workspace Business Starter is currently $7/user/month on an annual commitment before temporary promotions. Business email is separate from the website even when a hosting company bundles basic mailboxes.
Transactional email: Contact-form notifications, order confirmations, password resets, and application mail need reliable delivery. A dedicated transactional-email service or SMTP relay may be free at very low volume or add a small recurring cost at scale.
Photography, video, illustration, and fonts: Free stock assets can be enough for some sites. Custom photography or video is a separate production project and can materially change both the quality and the budget.
Copywriting and translation: Professional copy, editorial review, localization, and translation scale with content volume and the number of markets, not merely with development hours.
Analytics and consent management: Analytics tools may be free, but consent tooling, tagging implementation, server-side tracking, or a more rigorous measurement setup can add software and implementation costs.
Premium plugins, apps, and SaaS integrations: Booking systems, form tools, CRM products, search, subscriptions, ecommerce apps, and premium CMS extensions can all introduce recurring license costs.
Payment processing: In the US, Stripe’s standard domestic online-card rate is currently 2.9% + $0.30 per successful transaction. PayPal Checkout currently lists 3.49% plus a $0.49 fixed fee for US domestic PayPal Checkout transactions. Rates differ by payment method, country, plan, and cross-border usage.
CDN and edge services: Cloudflare has a free tier and currently lists Pro at $20/month when billed annually. Many sites never need to move beyond free or bundled CDN capability.
Backups and monitoring: These may be included in managed hosting or handled separately. The important question is not whether the line item exists, but who is responsible for recovery when something fails.
Accessibility testing: Accessible implementation belongs in the build. Formal audits, specialist testing, assistive-technology testing, or legal compliance work can add additional scope where required.
Content and SEO migration: Preserving URLs, redirects, metadata, internal links, structured data, and important content during a redesign is real work. Ignoring migration planning can damage an existing site’s search performance.
Training and handover: Someone needs to know how to publish, edit, recover access, and manage the accounts after launch.
Post-launch changes: Real users will expose things nobody noticed in staging. Agree in advance which defects are included in the launch warranty and which changes count as new scope.

How much should a small business actually spend?

There is no universal number, but there is a useful decision framework.

Start with the economic role of the website.

A site whose job is to show opening hours, establish credibility, and provide a phone number has a very different investment ceiling from a site responsible for generating a large share of the company’s qualified leads.

If the website is an acquisition channel, compare its performance with other channels. What does a qualified lead cost through paid advertising? What is a new customer worth? How much traffic reaches the site? If a business is spending heavily to send people to a weak website, improving the site can have a measurable payback period. If most work arrives through referrals and the website mainly confirms that the company exists, the case for a large custom budget is weaker.

Then consider the expected lifetime. A site expected to remain useful for several years can justify more up-front work than a campaign site that will be retired in six months.

Ask whether staff will use it operationally. A booking portal, content workflow, dealer area, or internal dashboard can save employee time as well as generate customer value. That changes the return calculation completely.

Finally, ask how costly it is for the website to be bad.

For a referral-driven company, a mediocre site may be mostly a credibility problem. For a business sending tens of thousands of dollars of paid traffic to it, poor conversion can waste money every day. For an ecommerce store, technical failures can directly block revenue. For a web application, failures can stop the product itself.

The right budget is therefore partly determined by the cost of under-investing, not just by what a supplier wants to charge.

When custom development is worth paying for

Custom development becomes easier to justify when one or more of these is true:

The website needs business logic or workflows that a standard platform cannot express cleanly.
Integrations with internal systems are important to the business.
Brand differentiation and conversion quality have meaningful commercial value.
The content model is complex enough that a generic page builder creates long-term editorial problems.
The site will evolve continuously and needs an architecture that can absorb change.
Source-code access, deployment flexibility, portability, or control over the infrastructure materially reduces business risk.
The cost of a mediocre implementation is high because the website is a major acquisition, ecommerce, or operational channel.

Custom does not mean “everything written from zero.” A sensible custom project still uses mature frameworks, libraries, hosted services, payment providers, CMS products, and infrastructure where they solve the problem well. The value is in controlling the architecture and experience around the requirements that actually differentiate the project.

When a website builder is genuinely enough

A single-location service business that needs a clean professional site quickly, with no custom workflows and only a handful of pages, may not need custom development at all.

A founder validating an idea usually should not spend five figures on a polished custom site before knowing whether the offer works. A temporary campaign page may be better built in the fastest tool the team can operate. A referral-driven business whose website mainly acts as a credibility check may get almost all of the value it needs from a strong template, good copy, and good photography.

Those are not second-class solutions. They are examples of matching the implementation to the economic job the site has to do.

The expensive mistake works in both directions. Underbuilding a website that is central to revenue can be costly. So can spending custom-development money on a site whose only requirement is to exist and look credible.

How to compare website quotes

This is where a lot of price confusion originates. The headline number is easy to compare. The assumptions underneath it are not.

A lower quote with vague scope is not necessarily cheaper. It is simply less defined.

Start with these questions.

What exactly is included, and how many unique templates does the quote assume? Get the number of unique layouts or components in writing. “20 pages” is not enough information.

Is the design based on an existing theme, a customized theme, or a bespoke design system? Those are different scopes and should be described honestly.

Who produces the content? Clarify copywriting, photography, video, translation, product data, legal text, and who is responsible for entering it into the CMS.

Who migrates the old website? If the site already receives search traffic, identify who owns URL mapping, redirects, metadata, content cleanup, and launch validation.

What CMS and editing experience will the team receive? Ask to see how a normal editor creates or changes content rather than only looking at the public design.

Who controls the domain, source code, hosting, repositories, analytics, and third-party accounts? Contractual ownership and operational control are not the same thing. You can own the work while still lacking direct access to critical infrastructure. Make sure the handover model is explicit.

Are premium licenses transferable? A theme, page builder, plugin, or SaaS account tied permanently to the supplier’s master account can become a migration problem later.

What is included after launch? Separate the defect-fix or warranty period from ongoing maintenance and new feature development.

How are scope changes handled? Every project discovers new information. The important thing is having a clear mechanism for accepting, rejecting, or pricing changes rather than relying on an undefined promise that “small changes” are included.

What recurring fees exist? Itemize hosting, maintenance, platform subscriptions, licenses, analytics products, consent tools, ecommerce apps, and support retainers.

What happens if the supplier relationship ends? You should know what can be transferred, what access you retain, what documentation exists, and whether another competent developer can continue the project.

Also ask whether accessibility, technical SEO, browser/device testing, backups, monitoring, analytics configuration, and deployment are part of the scope or merely assumptions. A well-scoped proposal should make those boundaries boringly clear.

A rough budget estimator

Website pricing does not reduce cleanly to a formula, so a useful estimator should classify complexity rather than pretend it can calculate a quotation to the dollar.

Start with the closest delivery category: DIY builder, professionally configured template, custom small-business site, custom marketing/content site, ecommerce, or custom web application. That gives you the base band.

Then move higher within the band as these become more substantial:

More unique templates or components than a basic project in the category would require
Professional content creation rather than client-supplied finished content
Bespoke design or brand work
Third-party integrations, weighted by the actual API and synchronization requirements
Genuine multilingual architecture and editorial workflow
Migration from an existing site with meaningful content or search equity
Custom accounts, data models, permissions, or workflow logic

The last item deserves special treatment. If accounts, databases, permissions, and business workflows are becoming central to the project, do not simply keep adding a percentage to a normal website estimate. Reclassify the project as a web application and scope the workflows directly.

You can walk through the same classification interactively with the website cost estimator. Its result should be treated as a planning range, not an automated quote.

FAQ

How much does a small business website cost?

For a professionally built small-business website, $3,000 to $18,000 is a useful 2026 planning band for many projects, with simpler template-based work below that and more complex marketing sites above it. Design depth, unique templates, content, integrations, and geography matter more than the size of the company itself.

How much does a 5-page website cost?

Page count alone is a poor pricing method, but five straightforward pages built from one or two reusable templates can reasonably fall in the $1,200 to $5,000 professional-template range. Five pages that each require distinct custom design, content, or functionality can cost several times more.

How much does it cost to hire someone to build a website?

Current Upwork web-developer data shows a $15 to $50/hour median range on its marketplace, with its own experience bands extending much higher for expert work. Agency pricing varies heavily by location and discipline. Clutch’s current web-design data, for example, lists roughly $100 to $149/hour in the US, Canada, and Australia, $50 to $99/hour in the UK, and $25 to $49/hour in several lower-cost agency markets.

Hourly rate is only one input. A more expensive developer who needs half as many hours can still produce the cheaper project.

How much does a website cost per month?

A simple builder site can cost roughly $20 to $70/month in platform fees depending on the product and tier. A professionally managed small-business site may run roughly $50 to $500/month for some combination of hosting, maintenance, support, backups, and software, depending on what is actually included. Ecommerce and web applications can run much higher. Payment-processing fees are separate.

How much does an ecommerce website cost?

A professionally delivered ecommerce site can start around $5,000, with a practical mid-range project often landing around $10,000 to $35,000. Projects involving ERP integration, complex catalogs, subscriptions, internationalization, wholesale logic, or custom checkout behavior can exceed $70,000.

The platform itself can be much cheaper than the implementation. Shopify Basic, for example, is currently $29/month when billed yearly in the US, before apps, implementation work, and payment processing.

Is WordPress cheaper than a custom website?

Not necessarily, because a well-engineered WordPress website can itself be custom development. What is usually cheaper is a theme-based WordPress implementation with limited bespoke work. A custom WordPress theme with structured content, integrations, tailored components, testing, and performance work belongs in the same commercial conversation as other custom websites.

Is Wix or Squarespace cheaper than hiring a developer?

Usually, yes, especially upfront. It can also remain cheaper for the full life of a straightforward site. It can become more expensive or restrictive if requirements later force substantial workarounds or a migration, but migration does not automatically erase the savings from several inexpensive years on the platform.

How much does website maintenance cost?

For a small-business website, $50 to $400/month is a useful planning range for many maintenance arrangements, but the word “maintenance” needs to be defined. A plan that only applies CMS and dependency updates is not equivalent to one that includes monitoring, backups, content changes, support response times, performance work, and development hours.

For custom applications, teams sometimes use roughly 15% to 25% of initial build cost per year as a planning heuristic for maintenance and incremental work. It is not a universal rule. Mature stable software can cost less to maintain, while an actively evolving product can easily spend far more because new development is being mixed into the maintenance budget.

How long does it take to build a professional website?

A straightforward template-based business site can often be delivered in roughly 2 to 6 weeks when content and approvals are ready. A custom marketing site with several unique templates, content work, and integrations commonly takes 6 to 12 weeks. A focused web-application first version may take 3 to 6 months or longer depending on workflow complexity, team size, integrations, testing, and compliance requirements.

Those are planning ranges, not promises. Content readiness and decision speed can affect calendar time just as much as coding effort.

Should I pay monthly or pay a one-time website development fee?

Hosted builders are naturally subscription products. Custom development is often easier to evaluate when the build cost is separated from the recurring operating cost: one clearly scoped project fee for creating the site, then a separate maintenance, hosting, or support agreement if needed.

A monthly website package is not inherently bad, but read the contract carefully. Understand the total cost over the expected lifetime, what happens when payments stop, who controls the accounts and source code, and whether the agreement includes actual ongoing work or simply spreads the original build cost over time.

Final thought

The useful outcome of a pricing guide is not a magic number. It is knowing enough to ask for the right scope.

If one quote is $4,000 and another is $18,000, do not start by asking which supplier is overpriced. Ask what each one thinks it is building: how many unique templates, what level of design, who is producing the content, what gets integrated, how the CMS works, what happens to the old URLs, who owns and controls the accounts, and what support exists after launch.

Once those assumptions are visible, large price differences usually become much easier to explain.

If you want to classify your own project first, the website cost estimator walks through the same framework. Or send me the actual scope, what the site needs to do, who is providing the content, and what it needs to integrate with. I’ll tell you what I think the project actually requires, including when the sensible answer is that you do not need custom development yet. Get in touch if that would be useful.

About the author

Written by Tiago Galvão

Full Stack Developer · Portugal | Switzerland

I've been building for the web since 2001. Full stack development across Vue, Nuxt, Astro, TypeScript, C#, .NET, and PostgreSQL, among others, with a habit of writing down what actually worked and what didn't once the dust settles.

More about me →