WordPress or custom code? If you are preparing to create your website, you have inevitably run into this debate. And you have probably noticed it turns into a turf war: WordPress agencies swear nothing else seriously exists, some developers swear WordPress is heresy. The two camps share one awkward trait: each recommends exactly the tool it sells.
I am an independent web developer in Hannut, and I have a particularity that makes my opinion a little more useful than average: I practise both. I built, repaired and administered WordPress sites for years, and today I develop most of my showcase websites in custom code. This choice of tool is, by the way, only one step of the project: if you are starting from zero, my complete guide to creating a professional website in Belgium puts this decision back into the whole journey, from the brief to going live.
In this article I hand you my honest decision grid: six concrete criteria, weighed one by one, with my stated preferences and the cases where I recommend the opposite of what suits me. By the end, you will know which tool matches your project, and above all why.
WordPress, CMS, custom website: what exactly are we talking about?
Let us first set the vocabulary, because the debate is often muddled by badly defined words. WordPress is a CMS, a content management system: free software you install on a hosting plan, providing a ready-to-use base with its administration interface, its graphic themes and its plugins. An enormous share of the web runs on it, from the personal blog to the large corporate site.
A custom website, in my mouth, is a website whose code is written specifically for the project: the structure, the styles, the features, everything is designed for that precise site, without a generic theme or a stack of plugins. Contrary to a common belief, this does not imply months of work: for a showcase site, it is often quicker to build cleanly than a WordPress properly configured, secured and slimmed down.
Between these two poles lies a whole gradient: WordPress with a purchased theme and thirty plugins, WordPress with a specifically developed theme, online builders like Wix, e-commerce platforms like Shopify. To keep this article readable, I compare the two formulas I am asked about most: the classic agency WordPress, theme plus plugins, and the custom-coded website. Shopify will get its section further down, because the online store is a case apart.
One last piece of context: I hold shares in nobody. WordPress fed me for years, custom code feeds me today, and I invoice both without qualms. My only bias is the result: the right tool is the one that serves your project, not the one that flatters the provider's habits.
Criterion 1: day-to-day autonomy for updates
First question to ask yourself, before even the price: what will you actually modify yourself on your website, and how often? Your honest answer to this question weighs more than all the technical arguments that follow.
WordPress built its reputation on this: its administration interface lets you add an article, replace a photo or create a page without touching the code. If you publish content every week, if you manage a diary, news, a team of writers, that autonomy has real value. It is the CMS's historic strong point, and I will not hide it from you.
The reality on the ground nuances this picture, though. For most of my independent and shopkeeper clients, the showcase website changes a few times a year: a price, an opening time, a photo of a completed job. For that rhythm, a full administration interface is a tank for fetching the bread: heavy, to be maintained, to be secured, and in the end almost never opened.
Many WordPress dashboards end up abandoned: the owner is afraid of breaking the layout, cannot find where a given text hides in the page builder, and ends up emailing their provider. The promised autonomy becomes theoretical, but its complexity, that stays invoiced and maintained.
My way of settling this criterion: give autonomy where it truly serves. On my custom websites, I integrate a targeted editable zone when a piece of content genuinely lives: the restaurant's menu, the news, the availability diary. It is the compromise I set up for my clients, from Hannut to the Namur region: your hands on what moves, without the weight of a full dashboard to tame.
Verdict: WordPress wins this criterion if you genuinely publish often, with several contributors. For a showcase site that evolves a few times a year, custom code with targeted editable zones is easier to live with day to day.
Criterion 2: performance and loading speed
Second criterion, and the first structural advantage of custom code: speed. A specifically coded website loads only what it needs, nothing else. A WordPress fitted with an all-purpose theme and a pile of plugins loads everything its components carry: script libraries, stylesheets for unused features, database queries on every visit.
It is not a fatality, let us be precise: a careful developer can make WordPress fast. But that demands permanent optimisation work, caching plugins, a regular hunt for the superfluous, and every new plugin puts the result back at stake. Custom code, on the other hand, is fast by construction: nothing to deactivate, nothing to compensate for, because nothing useless was added in the first place.
Why this criterion weighs so heavily: your visitors first, who leave without a second thought a website that drags, especially on mobile with an average connection, on the train or at the back of an industrial estate. Google next, which integrates the loading experience into its ranking signals. Your image finally: a slow website conveys the idea of a neglected business, exactly the opposite of what you are paying for.
If your current website crawls, I devoted an entire article to the solutions for speeding up a website that is too slow: you will find the most frequent causes and the concrete remedies, WordPress installations included.
Verdict without suspense: custom code wins, clearly. It is the main reason I favour it for showcase websites: for equal budget, the difference in responsiveness is visible to the naked eye, from the first visit.
Criterion 3: security and maintenance
WordPress is the favourite target of automated attacks. Not because it is badly designed, but because it is everywhere: robots roam the web permanently looking for installations where one plugin is not up to date. When a flaw is discovered in a popular plugin, entire batches of websites are compromised in series, without anyone targeting anyone in particular.
The practical consequence: a WordPress site demands real, regular maintenance. Updates of the core, the theme and every plugin, each time with the risk that one update breaks another. Tested backups, monitoring, patches. It is a genuine work item, often sold as a monthly subscription, and it is not optional: a WordPress left unattended is an open door.
The custom website presents a tiny attack surface in comparison. No administration interface exposed to the whole world, no third-party plugins of unknown code, no database to plunder for a simple showcase site. It does not demand weekly security updates, because it does not carry other people's code.
Let us be honest in the other direction: custom does not mean invulnerable. A website with forms, a client area or payment also demands rigour, whatever the tool. But the order of magnitude of the risk, and therefore of the maintenance budget, bears no comparison between a soberly coded showcase and an assembly of third-party components.
The question to ask before signing a WordPress quote
Who does the updates, how often, and what happens if a plugin breaks the website on a Sunday evening? If maintenance is neither included nor priced, the quote is incomplete: you will discover that cost later, at the worst moment.
Verdict: clear advantage to custom code for a showcase site. WordPress can be kept secure, but that security is a subscription of vigilance: legitimate, necessary, and to be budgeted from day one, not discovered after the incident.
Criterion 4: initial budget versus total cost
"WordPress is free." That is true for the software, and that is about it. The premium theme is paid for, the serious plugins are paid for in annual subscriptions, the configuration is paid for, and the maintenance is paid for every month. Opposite it, custom code sometimes displays a higher initial quote, but its cost of ownership is flat: the hosting, the domain name, and the evolutions you explicitly request.
The right calculation is therefore made over a website's real lifespan, that is, several years. Add to the initial quote the plugin subscriptions, the maintenance package, the emergency interventions after a failed update. The cheapest website at the start is rarely the cheapest at the finish, and this simple arithmetic is strangely absent from most quotes.
Let us break a stubborn cliché in passing: custom code is supposedly the luxury choice of the debate. With me, a custom showcase website starts at €380 and is delivered in 7 to 14 days, technical foundation and basic SEO included. It is no commercial feat: a soberly coded website, without theme licences or layers of plugins, simply costs less to produce and to keep alive than people imagine.
To situate all the formulas on the Belgian market, agencies, freelancers, online builders, I detailed the real budgets of a website in Belgium item by item: you will see there that the gap between WordPress and custom code is far smaller than announced, and sometimes reversed.
Verdict: an optical illusion at the initial quote, advantage to custom code on the total cost for a showcase site. For a big editorial site with many contributors, WordPress can regain the advantage: its fixed costs amortise when the tool is genuinely used every week.
Criterion 5: SEO and scalability
You read everywhere that WordPress is "good for SEO". The truth is flatter: no tool is good or bad for search rankings in itself. Google ranks pages, not software: it evaluates their content, their structure, their speed, their behaviour on mobile. A well-configured WordPress does the job. A well-built custom website does it too, and often better.
Why better? A question of precision. On a WordPress with a generic theme, part of the code escapes you: approximate markup inherited from the theme, scripts injected by the plugins, HTML bloated by the page builder. On a coded website, every tag, every piece of structured data, every internal link is exactly what the developer decided. That fine control, added to the speed of criterion 2, gives custom code a technical head start.
Scalability now, because this criterion is double. WordPress evolves by adding plugins: booking, multilingual, directory, member area, there is a module for almost everything, and it is a real strength when the need is standard. Custom code evolves by development: a little more effort for common needs, but no limit for specific needs, the ones that fit into no prefabricated box.
My field observation: the independents and SMEs of Wallonia rarely have exotic needs at launch. But many see arriving, along the way, THE particular need that makes their specificity: a quote calculator, an interactive rounds map, a trade-specific form. That is precisely where plugin-based websites pile up patches, and where custom code unrolls.
Verdict: a draw on basic SEO, advantage to custom code on technical finesse and specific evolutions, advantage to WordPress on standard evolutions at a very small budget.
Criterion 6: dependence on the provider
It is the sledgehammer argument of the CMS defenders: "with WordPress, you are not a prisoner of your developer, any agency can take over the site". The argument is serious, it directly targets my model as an independent, and I therefore want to treat it without dodging.
On paper, it is true: WordPress is known to countless providers, and your website does not depend on the survival of my business. In practice, taking over an agency WordPress is rarely the stroll announced: theme modified without documentation, proprietary page builder, a stack of plugins nobody knows which still serve. The newcomer diagnoses, sighs, and frequently proposes... to redo everything.
Dependence exists on the WordPress side too, simply displaced: you depend on a theme its author can abandon, on subscription plugins, on a page builder that locks your content into its format. Getting out often amounts to rebuilding. I have described elsewhere the signs showing that a redesign is due: several come straight from that stack become unmanageable, on websites that were not even old.
How I answer this risk on the custom side: standard, readable code, HTML, CSS, JavaScript, the languages every web developer practises, delivered with full access, hostable anywhere, and belonging to you entirely. A well-written custom website is taken over without drama. What imprisons is cobbled, undocumented code, and it exists in both worlds.
Verdict: a much more balanced criterion than the legend says. The real protection does not lie in the tool, but in three guarantees: ownership of your access, readability of what is delivered, and a provider who accepts in writing the idea that you might leave one day.
The false good reasons for choosing WordPress
Before giving you the cases where WordPress remains my sincere advice, let us clear out the arguments that come back in quotes and conversations, and that do not hold up.
"It is free." We saw it at criterion 4: the software is free, the website never is. Theme, plugins, configuration, maintenance: the best-marketed free product on the web turns into subscriptions by the second week. Comparing a "free" WordPress to a custom quote is comparing the price of a bare engine to the price of a car that drives.
"Everyone uses it." Popularity proves popularity, nothing else. An enormous share of the web runs on WordPress, including an enormous share of the slow, hacked or abandoned websites. This argument says strictly nothing about your project, your budget or your real need.
"You will be autonomous." Sometimes true, often theoretical, as seen at criterion 1. Real autonomy is measured after a few months: if the dashboard is never opened, you are paying for the complexity of a publishing tool to change two opening times a year.
"It is better for SEO." No: it is neutral, at best. What ranks a website is its content, its structure and its speed, not its software's logo. An SEO plugin makes up for neither an obese theme nor hollow texts.
"Everything can evolve." Everything, really? As long as the need matches an existing plugin, yes. The day your need is genuinely particular, the evolution is paid for in specific development... in an environment that was not designed for it. The scalability argument deserves to be turned over before signing.
The cases where I still recommend WordPress, and those where I advise Shopify
Here now are the situations where, in full knowledge, I advise a CMS rather than my own custom code. They exist, they are legitimate, and you may be in one of them.
You genuinely publish frequently. An active blog, a news website, an association or organisation communicating every week with several writers: WordPress's interface, its user roles and its publishing flow are designed exactly for that. Recreating all that tooling in custom code would make no economic sense, and I would tell you at the first meeting.
Your central need is standard and covered by a mature ecosystem. A member directory, a complex multilingual site, an online course platform: when a proven plugin already does the job at a reasonable cost, refusing it on principle would be stubbornness, not advice.
Your team already knows WordPress and genuinely uses it. You do not break a working workflow for the pleasure of technical purity. In that case, my role is to clean up the existing: lighten the theme, sort the plugins, secure, speed up.
And the online store? It is my third tool: to sell online, I most often recommend Shopify, rather than WooCommerce on WordPress or a fully custom development. The platform takes charge of hosting, security, payments and updates: for a merchant who wants to sell rather than administer, that peace of mind is well worth its subscription. With me, a store starts at €890, configuration and personalisation included.
The logic is the same in all three cases: the tool follows the need. High-performance showcase: custom code. Publishing machine: WordPress. Store: Shopify. Beware of the provider who has the same answer for all their clients: they are selling you their comfort, not your solution.
The honest-provider test
Ask them: "In which case would you advise me against your own tool?" A professional who masters their subject answers with precise examples. The one who answers "none" has just told you they only have a hammer, and that your project will be a nail.
How I decide for a new project: my method in practice
Concretely, when an independent or an SME contacts me, I never start with the tool. I start with three questions: what must the website accomplish, who will keep it alive, and with what overall budget over several years? The six criteria of this article flow from those three answers, and the tool falls out at the end of the reasoning, not at the beginning.
For the great majority of showcase websites of independents, tradesmen and professional practices, that reasoning leads to custom code: a few fast, solid, well-ranked pages, with an editable zone if a piece of content genuinely lives. It is the heart of my activity, from €380, delivered in 7 to 14 days, and it is a choice I stand by out of technical conviction as much as out of experience repairing WordPress sites.
For editorial projects and heavy standard needs, I say it frankly: WordPress, well assembled and well maintained. For selling online, Shopify in most cases. You will find the detail of my services, from website creation to the AI chatbot via automation, with public prices: transparency on prices is part of the advice.
A final word on migration, because I am often asked the question: moving from a tired WordPress to a custom website is done without losing your rankings, provided the method is respected. Keep the addresses of the pages that rank or redirect the old ones cleanly to the new, preserve the content that works, check indexing after the switch. It is a marked-out process, not a leap into the void.
What to remember from this whole comparison
Judge on six criteria, not on the tool's fame. Count the total cost over several years, maintenance included. Demand ownership of your access and your code whatever the choice. And flee anyone who recommends the same tool to everyone.
My final recommendation: put the six criteria in writing, note your situation against each, and add it up. Real autonomy, speed, security, total cost, SEO and evolutions, dependence. In most cases, the answer becomes obvious before even asking for a quote. And if it does not, that is precisely the conversation I propose we have: frank, costed, and without tool religion.
