Not every SaaS needs a separate React app. Here is how I decide between Blade with Livewire, Inertia and a standalone SPA.
About half the founders who contact me have already decided their product needs a React front end and a separate API. Sometimes they are right. Often they are paying for two codebases, two deploys and an auth layer they do not need yet.
When Blade or Livewire is enough
If the product is mostly forms, tables and settings pages used by a few hundred business customers, server-rendered pages are faster to build and easier to change. A CRUD-heavy back office I built in Livewire for a Dublin logistics startup took seven weeks. The same scope as an API plus React app was quoted to them at fourteen.
When Inertia is the sweet spot
When the team wants React components, rich interactions like drag and drop boards or live charts, but there is only one client: the web app. Inertia keeps Laravel routing, validation and auth, and lets the front end be proper React. It is my default for new SaaS products in 2026.
When a real API pays for itself
A separate API makes sense when there is a mobile app, public integrations for customers, or more than one front end on the roadmap within a year. Then I build a versioned JSON API with Sanctum tokens, API resources for every response and a contract test suite, and the React app is just the first client.
The cost of getting this wrong is not the first build. It is the second year, when every small feature means changing a migration, a resource, a TypeScript type and a component, and the team has three people.
Comments