Next.js to Webflow

Migrate from Next.js to Webflow

This one is rarely about the code. The marketing site lives in the same repo as the product, every headline change is a pull request, and the people who write the copy cannot ship it. I move the marketing pages to Webflow and leave the app exactly where it is.

Book a call

A migration like this usually takes 3–4 weeks

What the migration covers

What moves, and what needs a decision.

Carried across

Marketing pages rebuilt in Webflow, matching the current design
Blog and any MDX content into CMS collections
Metadata, OG images and structured data rebuilt natively
URLs preserved exactly, so nothing has to be redirected that does not have to change
The app left alone, still deployed the way it is today

Flagged upfront

One domain across two platforms means a proxy: the app keeps the domain and rewrites the marketing paths to Webflow, since it cannot work the other way round
On Vercel every rewrite, redirect and header rule shares one routing budget, framework-generated routes included, so the marketing paths are proxied in groups rather than one by one
Webflow has no equivalent of an OG image generated per request, so those become real images or a build step
Anything the marketing pages import from the app stops being shared the day they split, and Webflow has no way to consume a design system from a repo

The reason to do this is almost never performance and almost always ownership. A marketing team that has to open a pull request to change a headline does not run experiments, and the site stops being a marketing surface and becomes a release artefact.

The part that needs deciding first is the domain. Both platforms want to serve example.com, and only one can. In practice the app keeps the domain and rewrites the marketing paths through to Webflow, because Vercel can rewrite to an external address and Webflow cannot. On Vercel that routing sits in one budget shared by every rewrite, redirect and header rule on the deployment, including the ones your framework generates, which is worth knowing before designing a hundred marketing routes.

The honest losses are the dynamic parts. An OG image rendered per request has no Webflow equivalent, so it becomes a real image or a generation step. Internationalised routing, incremental regeneration and anything reading request headers do not come across either. None of that is usually load-bearing on a marketing site, but it is better found now than in week three.

What you get back is a site the marketing team can change on a Tuesday afternoon, and an app repo that no longer carries a landing page directory nobody on the engineering side wanted to own.

Other migrations

Coming from somewhere else?

Reach out and see if we are a good fit.

Start a conversation

Currently booking two to four weeks out.