At Takaneo, the technology behind a website is chosen after one or more meetings with the client, never from a catalogue and never on the basis of how popular a platform happens to be.
That first conversation is there to listen to your objectives and to establish the practical requirements: who will produce the content, how often, for which users, on what recurring budget and with what expected lifespan. When the meeting is over, the agency does not come back with a list of options but with one or two technologies, argued in front of the client: what each one makes possible, what it costs over time, and why the others were set aside. The client then decides with full knowledge of the trade-offs, and the site answers real needs rather than the habits of its supplier.
Four cases cover the majority of projects:
Part of the confusion comes from how lopsided the market is. According to W3Techs figures published on 11 August 2026, WordPress powers 41.1 % of the websites measured, against 5.3 % for Shopify and 0.7 % for Drupal. A business owner asking their network, or an AI assistant, will therefore always get the same kind of answer, whatever the project.
The second source of confusion is the question itself. The four options are not comparable: WordPress and Drupal are content management systems, Shopify is a hosted commerce platform, custom development is a method rather than a product. The choice is not between four comparable products, it is between four ways of working.
In Luxembourg one constraint is almost always added to the list: multilingualism. Three or four language versions of the same site, maintained over the years by a small in-house team, weigh more heavily on the technical decision than any feature shown in a demo. That is also why Takaneo starts with a listening session rather than a technical answer: a platform chosen before the objectives have been heard produces a site nobody updates a year later.
For a company that presents its services and publishes news or expert articles, WordPress does the job without argument. The editing interface can be learned in half a day, and the entry cost remains the lowest of the four options.
Several sites built by Takaneo fall into this category. Maison Bosk presents its wood flooring collections. Trendy Stone showcases its decorative floor finishes, polished concrete, terrazzo and resins. Fondatioun Kriibskrank Kanner publishes its news, its events and its donation appeals in French and Luxembourgish, as does Île aux Clowns . In each case a small team had to be able to publish on its own, without going through a developer.
Its greatest strength is also its limit, the plugin. A healthy WordPress runs on a handful of carefully chosen and maintained plugins. One that accumulates thirty of them over the years, installed by a succession of suppliers, becomes slow, fragile and expensive to bring back up to standard.
Two signals indicate that WordPress is no longer the right fit:
Drupal does not address the same client, and its minority market share says nothing about its relevance. It becomes the right tool when the site is produced by an organisation rather than by a person: a public institution, a federation, a multi-entity group, several departments each publishing their own scope.
What it brings comes down to three points:
The price is real: a higher entry cost, scarcer profiles, and version upgrades that need proper planning.
Once selling online becomes the heart of the project, the question of the CMS matters less than the reliability of the checkout, the payment methods, stock management and logistics. Shopify handles all of that by default, updates itself without intervention and absorbs traffic peaks, which a shop grafted onto a general purpose CMS does not guarantee.
Fearite , a young brand of products designed to help children deal with their emotions, whose catalogue is organised by range and by emotion, runs on Shopify for exactly that reason: cart, payments, stock management and shipping are handled by the platform, and the team spends its time on its products rather than on the machinery of selling.
The Luxembourg market remains lightly equipped in this area. According to Eurostat data extracted in February 2026, 14.7 % of Luxembourg companies made online sales in 2024, against an average of 23.6 % across the European Union. Exceptions do exist, such as a catalogue with very particular pricing logic or a deep integration with an ERP, and they justify looking at another route.
Custom becomes relevant when the value of the project no longer sits in the pages but in what happens behind them: client area, simulator, booking, constant exchanges with business software. Starting from a CMS then means spending half the budget working around what it imposes.
The Institut National du Cancer website follows this logic: three languages, a patient area and a professional area with distinct journeys, an events calendar, a map of the Luxembourg oncology ecosystem and a separate information portal for the general public, a combination no off the shelf theme covered.
The real deciding factor is not feasibility but handover. A custom site lives for as long as a team can take it over, which requires documented code, up to date dependencies and intellectual property that is clearly assigned. Without those conditions, custom development creates a stronger dependency than any proprietary platform.
There is also a middle road, often the fairest one: a standard foundation for the content, extended by development limited to the functions that carry the value.
The budget for a site does not stop at launch, and that is where the technology choice is paid for. The State of WordPress Security report published by Patchstack on 25 February 2026 counts 11,334 vulnerabilities identified in the WordPress ecosystem during 2025, up 42 % year on year, of which 91 % come from plugins rather than from the core software. The figure is not a condemnation of WordPress, which remains the most heavily audited CMS in the world, but it puts a number on the cost of a site left unattended, hence the importance of regular maintenance, a subject Takaneo will return to in a future article.
Based on its years of experience, Takaneo has changed the way it quotes: we now systematically add a maintenance line to our proposals, presented as an option, so that no cost appears after the site goes live. An unmaintained site always costs more than a service contract, because the catch up work is paid for in a hurry, usually after an incident.
W3Techs, Usage statistics of content management systems, 11 August 2026, https://w3techs.com/technologies/overview/content_management
Patchstack, State of WordPress Security in 2026, 25 February 2026, https://patchstack.com/whitepaper/state-of-wordpress-security-in-2026/
Eurostat, E-commerce statistics, data extracted in February 2026, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=E-commerce_statistics
Would you like a website that genuinely serves your business?
Request your 30 minutes of free advice by clicking here!
Aurélien Sibret
Digital officer at Takaneo, in charge of web projects and digital production
The decision is made in a meeting, not in front of a comparison table. A first appointment is used to understand the business, the users it serves, how content will be produced and which tools are already in use internally, from invoicing software to booking systems. A second one confronts expectations with the real budget, separating what is essential at launch from what can wait for a second version. At Takaneo the technical recommendation comes after those conversations, written down and argued, including the reasons the other options were set aside. An agency that names the technology before listening is selling its own habits.
The cost of a site depends far more on its functional scope and the volume of content to produce than on the technology chosen. The same showcase site on WordPress or on Drupal differs mostly in initial configuration time and in the profile of the people who maintain it afterwards. Shopify introduces a different logic, with a monthly subscription and a commission on transactions, which shifts part of the budget from investment to running costs. Custom development concentrates the cost at the start and rarely reduces it later. Takaneo recommends budgeting three full years, covering build, hosting, maintenance and improvements, rather than the launch invoice alone.
Changing platform is always possible, but the cost depends entirely on how the content was structured in the first place. A site whose content sits in distinct, consistent fields migrates cleanly and often automatically. A site where everything was composed in free visual blocks has to be migrated by hand, page by page. That is why structuring content deserves more attention than choosing the tool itself. Sound modelling at the outset guarantees that the technology question can be reopened later without starting from scratch.
WordPress is not inherently less secure, it is simply far more exposed. Its dominant position makes it the preferred target of automated attacks, and its open plugin ecosystem multiplies the potential points of entry. The Patchstack report of February 2026 shows that 91 % of the vulnerabilities recorded in 2025 came from plugins and themes rather than from the core software. A site with few plugins, updated regularly, backed up and monitored carries a risk comparable to any other solution. The deciding factor is maintenance, not the brand of CMS.
A hosted solution such as Shopify takes care of updates, security and availability, in exchange for a subscription and a framed level of customisation. Installed software such as WordPress or Drupal leaves full control over the code, the hosting and the data, but transfers responsibility for upkeep to the site owner or their agency. The choice depends on how much control is genuinely needed and on the company's capacity to take on that follow up. An organisation with no internal technical resource and no maintenance contract is often better served by a hosted solution.
Multilingualism has to be decided before the technical choice, because it determines the workload for years to come. The real question is not whether the platform can handle several languages, they all can, but who will write and proofread each version, and what happens when one language falls behind the others. Machine translation left unchecked damages credibility as much as readability. Takaneo advises defining a reference language, a clear rule for untranslated content, and a translation budget booked as a running cost rather than as part of the initial project.
Search engine optimisation does not depend on the technology used. Search engines assess the quality and structure of the content, page speed, the clarity of the markup and the coherence of internal links, all of which WordPress, Drupal, Shopify and custom development can handle correctly. A badly designed custom site ranks lower than a rigorous WordPress one. The only real advantage of custom development is the absence of superfluous code, which makes performance work easier. That advantage disappears the moment the published content is not up to standard.
