Around 100 websites, landing pages and relaunches. A system is not an end in itself: what matters is that editorial work, store, ERP and visibility play together afterwards instead of working against each other.
Most website projects do not fail on the technology. They fail because nobody can maintain the result, because the product data comes from a source nobody thought about, or because the new site is prettier than the old one and still produces fewer enquiries.
So the starting point is not a system comparison but four questions: who maintains the content, and how technical is that person? Where do product and price data come from, and how often do they change? What has to be connected, ERP, CRM, shipping, accounting? And which searches should bring customers here in the first place?
Only once those four answers exist does the choice of CMS or store platform get made. In practice that rules out half the options by itself, and the remaining choice is no longer a matter of taste.
The right call when several countries, several languages and several editors come together and permissions have to be cleanly separated. For cetrac.com: TYPO3 12, a new information architecture, category and product structure, Solr-backed search in the store, seven language versions delivered single-handedly. TYPO3 costs more to set up and only pays off above a certain size.
The fastest route to a site an editorial team maintains without technical knowledge. For stores up to medium complexity with WooCommerce. What matters is avoiding the plugin sprawl that makes WordPress sites slow and vulnerable after two years: a few well-maintained extensions instead of twenty half-used ones.
Stores meant to sell rather than merely exist. Shopify when it needs to be fast and operations should stay simple. Shopware when product logic, price tiers or B2B requirements get more complex. Either way: a store is only as good as the data flowing into it.
Also JTL, Webflow, Joomla, Drupal, Contao, Wix, Squarespace and headless setups. JTL when the ERP leads. Webflow when design is in front and there is little editorial work. Headless when several channels are served from one source.
A new information architecture, a complete URL inventory, old-to-new mapping with 301 redirects, transfer of the content that actually earns clicks, consistent hreflang and monitoring in the weeks afterwards. A relaunch must not cost visibility, and that is exactly what happens alarmingly often.
Market adaptation instead of literal translation, dedicated keyword research per target market, consistent hreflang. Seven languages for cetrac.com, seven for the EasyFlow Suite including Hebrew with right-to-left layout. I work in Spanish, German and English myself.
The four systems people ask about most have a page each. They do not only say what the system is good for, but also when I advise against it.
The fastest route to a site an editorial team maintains itself. Without the plugin sprawl that makes WordPress sites slow after two years.
Go to WordPressFor ranges with many variants, customer group prices and a leading ERP system. The effort sits in the data, not the design.
Go to ShopwareWhen nobody in-house should look after a store technically. The platform handles hosting, security and updates.
Go to ShopifyWhen several countries, languages and editors come together. More effort to set up, but sustainable over years.
Go to TYPO3A relaunch is the moment a site can lose visibility built up over years without anyone noticing immediately. The new site looks better, everyone is pleased, and three months later someone asks why the enquiries have dropped.
The cause is almost always the same: nobody recorded which URLs existed and which of them earned clicks. The new site has different addresses, the old ones lead nowhere, and Google starts over on the affected pages.
Record every existing URL with clicks, impressions and positions from Search Console. Only that list shows which pages actually perform.
Every old URL gets a destination on the new site. What disappears without replacement is a deliberate decision, not an oversight.
Permanent redirects, and no blanket redirect of every legacy URL to the home page, because Google treats that as an error.
Text that ranks gets carried over and improved, not cut for design reasons.
Watch indexing, errors and positions in the first weeks, so a problem shows up while it is still small.
That is decided not by the technology but by the team who will maintain the site afterwards. If there is an editorial team with several people and several languages, TYPO3 makes sense. If one person occasionally updates text, WordPress gets there faster. If selling is the point, product complexity decides between Shopify, Shopware, WooCommerce and JTL. So I ask first about editorial capacity, product data and interfaces, and only then about the system.
A website builder is cheap and fast as long as nothing special is needed and nobody is responsible for visibility. An agency brings a team and processes, at the cost of alignment loops and overhead. In-house delivery is fastest but reaches its limit on projects that need several disciplines in parallel. Which route holds up depends on who maintains the site afterwards.
Not if someone records the old URLs beforehand. That is exactly where it usually goes wrong. A relaunch needs a complete URL inventory, an old-to-new mapping with 301 redirects, transfer of the content that actually earns clicks, and monitoring in the weeks afterwards. The cetrac.com relaunch rebuilt the entire category and product structure this way.
Yes. A migration is two thirds data work and one third technology: clean up the article master data, reorder variants and categories, map images and attributes, and only then move. Anyone who takes the data across unexamined ends up with the old situation in a new system.
Yes. For cetrac.com I set up seven language versions single-handedly, and for the EasyFlow Suite seven languages including Hebrew with right-to-left layout. Instead of literal translation the approach is market adaptation, with dedicated keyword research per target market and consistent hreflang. I work in Spanish, German and English myself.
That is the goal. A site that needs an agency for every text change will inevitably go stale. So the editorial structure is built so it can be maintained without technical knowledge, and on request I train the people who will work with it later.
Not sure which system fits your situation? Ask me, I am happy to sort it out, even if the answer is a different system.