When to leave Webflow

Most posts about leaving Webflow are written by people who sell the thing you would leave for. This one is written by someone who sells the thing you would be leaving, so weigh it accordingly, but I have moved sites in both directions, and the Webflow to Astro page exists because that work is real.

Here is what I have actually seen. Roughly half the companies that decide to leave Webflow do not have a Webflow problem.

The signals that hold up

You have hit a cap that does not move. Not "the CMS is getting full", but genuinely at the ceiling, on the highest plan that makes sense, with content that is still growing. The CMS limits post covers how to tell the difference. This is the cleanest reason to go, and the rarest.

The site needs to decide something before it is sent. Pricing that depends on who is asking, content gated for real rather than for show, region-specific legal copy. Webflow renders the same page to everyone and then lets JavaScript change it, which is fine for personalization and not fine for anything that must be true before the HTML leaves the server.

A developer is now in every change. This is the interesting one, because it is a reversal. The reason to be on Webflow is that marketing ships without engineering. If enough custom code has accumulated that every change routes through a developer anyway, you are paying platform constraints for a benefit you no longer get.

The site is one part of a product, not a site in front of one. When the marketing pages and the application want to share components, routing, auth and a design system, keeping them in two systems costs more than consolidating.

The signals that are really a bad build

"The site is slow." Almost never the platform. Webflow serves from a CDN and the sites I audit are slow because of third-party scripts and uncompressed hero media. Moving platforms carries those with you, and a site that was slow on Webflow is slow on Next.js with the same tag manager on it.

"The Designer is a mess." That is a class-naming problem, and it followed a person out of the door. A rebuild fixes it. A migration fixes it too, expensively, while introducing a new set of things nobody has named yet.

"We have too much custom code." Worth counting before concluding. Three embeds and a script is not too much custom code; it is a normal site. The threshold is whether a non-developer can still change the things they need to.

What moving actually costs

The number people budget is the rebuild. The numbers they do not are the ones that hurt.

Every URL has to be mapped and redirected, or the rankings that took two years go. The content has to come out, and an export is not a website: it is fields in a shape that suits the old system. Whatever your marketing team could do alone, they now cannot, until someone builds them a way to do it. And for some period you are running two things.

That is not an argument against leaving. It is an argument for being sure, because the cost is front-loaded and the benefit is not.

The option most people skip

You do not have to move all of it.

The split that works more often than a full migration: the marketing site stays on Webflow, where marketing can keep shipping, and the part that outgrew it (the application, the large data set, the logic that has to run on a server) moves to something built for it. The two meet at a seam you design on purpose.

This is the same architecture I argue for on the marketplace and EdTech pages, and it is not a compromise. It is what most companies that "left Webflow" actually did, described accurately.

If you are weighing this and want a second opinion from someone with an obvious bias, say so directly: I will tell you when the answer is that you should go.