Custom website vs template: what speed and ownership really cost

When a business commissions a new website, the choice is often framed as custom code versus a template. It is really a decision about how much of the site is shaped around the business, how much unused machinery it carries, and how easily it can become part of the sales process.

A template can be exactly the right tool. A custom build can also be the wrong tool if the business only needs a simple presence. The useful question is what the site needs to do now, what it may need to do next, and what visitors experience on real devices.

A template is a decision somebody else made about your business

A template is not simply a visual shortcut. It comes with assumptions about navigation, page structure, content hierarchy, components and editing.

That can be useful because common problems such as responsive layouts, contact pages and product grids have already been solved. But those decisions can become constraints when the business has an unusual customer journey or needs integrations.

Imagine a service company where the main conversion is not “read more” but “tell us about your project, qualify it, route it to the right person and get a call booked”. Or an online shop with complex product rules and a specific fulfilment workflow. The template may display the content, but the business can end up adapting its process to the shape of the site.

That is the central difference in the custom website vs template decision. With a template, you adapt the business to a predefined system. With custom development, you can define the system around the business. Neither is automatically better.

For WordPress, themes are collections of files that control presentation and can include template files, stylesheets, images and code. WordPress also allows themes, widgets and plugins to extend what the site can do. WordPress explains how themes work.

Where templates are genuinely the right call — say this early and mean it

If you need a five-page brochure site, have a limited budget, and mainly need clear information and contact details, a good template can be entirely sensible. You do not need bespoke architecture just to publish an About page, Services page, Contact page and a few pieces of content.

The same applies to a shop that genuinely needs to be live next week. A mature hosted store platform or established e-commerce theme can provide a catalogue, checkout, payments and basic administration without requiring every component to be built from scratch.

Templates also make sense when the business model is conventional and unlikely to change materially. Standardisation can reduce the amount of custom code the team has to understand. For a more complex shop, custom e-commerce development becomes more relevant when catalogue rules, checkout, fulfilment or integrations stop fitting the standard model.

The mistake is treating “template” as synonymous with “bad” or “custom” as synonymous with “high quality”. A poorly built custom site can be slow and hard to maintain. A carefully selected template can be quick, stable and adequate.

What a page builder actually ships: the weight of every feature you did not use

A page builder gives non-developers a large set of components and controls. The trade-off is that the site is being built inside a general-purpose system. Depending on the builder, theme and plugins, the browser may process code, styles, assets and third-party requests supporting capabilities that a particular page never uses.

That does not mean every page builder site is slow. Caching, image optimisation, sensible plugins, capable hosting and careful loading strategies can produce a fast result. Custom code can also be poorly optimised.

Page-builder flexibility can mean carrying more layers than a particular page needs. The useful comparison is the delivered experience, not the builder name.

The practical issue is what those extra layers mean later. Dependencies can need updating, testing, loading or debugging. The page builder site speed question is therefore not “is this builder fast?” but “what does this particular page actually ship?”

A custom build can be deliberately narrow: if a component is not part of the requirement, it does not have to be part of the implementation.

This is also where “WordPress template limitations” are often misunderstood. The limitation may not be that WordPress cannot do something. It may be that achieving it inside a chosen theme and builder requires additional layers or compromises.

Speed is a business number, not a Lighthouse score

A performance score is useful for finding problems. It is not the number that pays the bills. The commercial question is what happens to people who arrive from search, advertising, email or a referral and encounter a slow page.

Google and Deloitte's “Milliseconds Make Millions” study analysed mobile data from 37 brands across retail, travel, luxury and lead generation. In its retail analysis, a 0.1-second improvement across four site-speed metrics was associated with an 8.4% increase in transactions and a 9.2% increase in average order value. Those are observations from that study, not a universal conversion formula.

Portent's research, based on more than 100 million page views across 20 B2B and B2C sites, found that B2B sites loading in one second had a conversion rate three times that of sites loading in five seconds. For B2C e-commerce, its analysis found a 2.5-times difference between one-second and five-second load times. These are findings from a particular dataset.

Akamai reports that its research has found a 100-millisecond delay can reduce conversion rates by up to 7%. Its earlier retail research also reported that 53% of mobile visitors would leave a page taking longer than three seconds to load. These figures come from Akamai's research and should be read in that context.

So what does speed cost in money? Start with your own funnel. If 10,000 people reach a product or enquiry page, the value of a one-percentage-point conversion gap is simply the number of lost outcomes multiplied by the value of those outcomes. Your analytics should tell you that number.

Core Web Vitals and conversion should be kept related but distinct. Core Web Vitals help diagnose the experience. Google's current set covers loading performance through LCP, responsiveness through INP and visual stability through CLS. Google's Web Vitals guidance recommends measuring at the 75th percentile, with good thresholds of 2.5 seconds or less for LCP, 200 milliseconds or less for INP and 0.1 or less for CLS. They are useful experience targets, not a substitute for revenue analysis.

Measure on the device that decides: a mid-range phone on a real mobile network, not a developer's laptop on office wi-fi

A website can feel instant on a developer's laptop and still feel slow to a customer. Browser capability, CPU, connection, location, caching, image size and JavaScript all affect what the visitor experiences.

Test important pages on a representative mid-range phone and a real mobile connection. Test the homepage, landing pages, product or service pages, forms and checkout, not only the page that scores best. Google's Web Vitals guidance also points site owners towards field measurement and real-world user experience.

The part most comparisons miss: a site that cannot hand a lead to something that calls them is a brochure with a form stapled on

A form is not the end of a lead-generation process. It is the point where a website should hand useful information into the next system.

For a service business, that might mean capturing the enquiry, identifying the source, routing it to the right person and triggering a prompt response. The question is not simply “does the form submit?” but “what happens next?”

The website is more useful when the enquiry continues into the sales process rather than stopping at a form submission.

A custom build can be designed around that hand-off rather than adding an integration after the brochure is finished. The workflow might include a CRM, email automation, calendar, notifications or an AI voice agent that responds to a new lead.

If speed-to-lead is important, the website and response system should be briefed together. BlueX's outbound speed-to-lead service is an example of treating the website enquiry as the start of the workflow rather than the end of the website.

Ownership and lock-in — who holds the code, what happens when you stop paying a platform, and what “you can edit it yourself” should and should not mean

Before signing off a build, ask what you actually receive. The agreement should make clear who owns or can use the custom code, where the repository lives, who controls the domain, hosting and key accounts, and how the site can be moved if the relationship ends.

Hosted platforms need a different question. If the site depends on a proprietary service, stopping that service can affect the hosted storefront or editing environment. What can be exported, what must be rebuilt and what remains portable depends on the platform and its terms.

“You can edit it yourself” should mean that your team can change the things that change regularly: products, prices, blog posts, images and selected page content. Those areas should have a sensible admin interface.

It does not mean every part of the website should be exposed as editable blocks. Layout systems, business logic and performance-critical code can stay as code. Stable structure can remain controlled implementation while the frequently changing content stays accessible to marketing.

A short checklist for briefing whoever builds it

• Start with the business process. List the actions the website must drive, not just the pages it must contain.

• Separate fixed and editable content. Decide what marketing staff should change without a developer.

• Ask what is custom and what is inherited. Get a clear explanation of the theme, builder, plugins, integrations and custom code.

• Make performance a requirement. Specify mobile testing, realistic network conditions and the pages that matter most to conversion.

• Measure commercial outcomes. Track enquiries, purchases, calls and booked meetings, not only performance scores.

• Map the lead hand-off. Know exactly where a form submission goes and what happens next.

• Clarify ownership. Confirm access to the domain, hosting, source code, analytics, CMS and third-party accounts.

• Plan for change. Ask what happens when the catalogue, service model, integrations or traffic grow.

If you are briefing a team on a site that needs to sit between marketing, e-commerce and lead handling, BlueX's website and e-commerce development service is one place to start. BlueX also covers the wider digital work through its full range of services. If you already know what you need, you can contact BlueX with the brief.

FAQ

Is a custom website always faster than a template?

No. Custom code gives a team more control over what is shipped, but poor implementation can still be slow. A well-built template site can be fast when its theme, plugins, assets and hosting are controlled.

Why is my website slow if it does not have many pages?

Page count is not the main measure of browser work. One page can still load large images, scripts, fonts, third-party services and builder components.

Do good Core Web Vitals guarantee better conversion?

No. Core Web Vitals describe important parts of user experience, while conversion also depends on the offer, message, trust, product, traffic quality and sales process.

When should an e-commerce business consider custom development?

Consider it when standard catalogue and checkout behaviour no longer matches the business process, or when important workflows need unusual integrations or rules. If the shop is conventional and needs to launch quickly, a mature platform or theme may be more practical.

What should I ask an agency before choosing between a template and custom build?

Ask them to show how the proposed stack handles your actual customer journey, not just the visual design. Also ask what you can edit, what remains code, how performance will be tested, where lead or order data goes and what happens if you later move providers.