Why Split Backend and Frontend

Diving deeper into

Thom Krupa, co-founder of Bejamas, on building dynamic apps on the Jamstack

Interview
it's split backend and frontend
Analyzed 6 sources

The real payoff of splitting backend and frontend is that a website stops behaving like one fragile machine and starts behaving like interchangeable parts. Content lives in a headless CMS or other API service, while pages are built and served from a CDN through platforms like Netlify or Vercel. That lets teams swap pieces independently, ship frontend changes faster, and handle traffic spikes without scaling a monolithic app server.

  • In a monolith like WordPress, the editor, database, templates, and web server are tied together. In Jamstack, the CMS just returns data over an API, and the frontend turns that data into static or server rendered pages. That separation removes much of the plugin, patching, and server maintenance burden.
  • The split also changes who can build what. Frontend developers can use managed backend pieces like auth, search, payments, or databases without owning a full custom backend, which is a big reason Vercel and Netlify became popular as push button deployment layers on top of commodity cloud infrastructure.
  • This architecture makes the stack more modular for businesses. A team can keep WordPress for editors, Shopify for commerce, and a custom React or Next.js frontend for the customer experience, instead of forcing every workflow into one system’s built in templates and plugins.

The direction of travel is toward even thinner separation, where static delivery, serverless functions, and edge execution blur together. The frontend will keep acting as the control layer, while backend capabilities become more like plug in services behind APIs, making the split more powerful rather than less important.