WordPress or a custom build?

The question usually comes up when a project is a bit bigger than an ordinary website but not yet clearly an application. The answer doesn't come from a preference in technology. It comes from two practical questions: who manages the content and how unusual is the business logic.

What WordPress does well#

WordPress is a content management system. It's good at what it was built for: letting a non-engineer create, edit and publish structured content.

A custom theme with ACF blocks gives you a situation where a marketing person assembles a new campaign page from existing blocks without breaking the design. That's WordPress's strongest suit, and the reason it's the right choice for most company websites.

On top of that: with WooCommerce, the local payment methods, parcel machine services and accounting connections already exist. Rewriting those for a custom store is a cost that rarely pays for itself.

Where WordPress starts to push back#

WordPress is built around documents - posts and pages. The further your data model moves from that, the more you end up fighting it.

Signs you're holding on to the wrong tool:

• users log in and each role sees a completely different view
• objects have states and approval steps, not just "draft / published"
• there are complex relationships between records that you're keeping in post meta fields
• every new feature adds a plugin, and every plugin adds a conflict
• most of your traffic isn't reading content, it's doing something

At that point the problem isn't WordPress. The problem is that you're building an application inside a content management system.

A practical decision tree#

1. Is the main use reading content? Yes → WordPress. No → keep going.

2. Do users have to log in and enter data? No → WordPress. Yes → keep going.

3. Are there more than two roles, each seeing what amounts to a different application? No → WordPress with a custom build is probably enough. Yes → keep going.

4. Does the business logic have states, permissions and rules that need to be testable? Yes → a custom build.

The most common right answer isn't either extreme. It's both: a public website in WordPress that marketing manages themselves, and a separate application behind it that developers work on. The two talk over an API and neither gets in the other's way.

What a custom build actually costs#

The cost of a custom build isn't only the development. It also includes the parts that come free with WordPress:

• user management, permissions and password resets
• an admin view for managing content
• a media library and image processing
• forms and validation
• the SEO basics, sitemap, metadata
• security updates that somebody has to keep watching

Once those are counted, the starting price of a custom build is considerably higher. It pays off when the business logic is distinctive enough that bending off-the-shelf parts would cost more.

And which technology, then?#

If the answer is "custom", we choose the stack to fit the job. We've built on Laravel and Vue, and on custom PHP. If you already have a development team or a preferred stack, we work with it - the result has to stay something you can maintain, not something that depends on whoever built it.

The wrong answer in either case is the one a developer picks based on what they'd like to learn next.