How to migrate from WordPress to Webflow

Most guides to this migration are written by people who sell the destination, which is why the export and import steps get four hundred words each and the part where your event calendar plugin turns out to have no equivalent gets nothing.

So this page does both halves. The losses come first, because they decide whether you should start at all, and the walkthrough follows in the order you will actually work through it.

Should you migrate at all

There are sites where the honest answer is no, and it is worth checking whether yours is one of them before you start pricing anything.

Stay on WordPress if your site has logged-in users. Memberships, course platforms, forums, gated content and customer accounts are the clearest case. These are not features you rebuild in a weekend, and they are usually the thing the site exists for.

Stay if you are running WooCommerce and it works. Moving a functioning store to a different commerce platform is a project with its own risk profile, and bundling it into a site redesign is how both go wrong at once.

Stay if the site's value sits inside a plugin with no equivalent. Booking systems, directories, calculators, anything with its own database tables. You are not migrating a website at that point, you are commissioning a replacement for a piece of software.

Stay if nobody has budget for a rebuild. This is the one people talk themselves out of. A migration is not a port, and treating it as a data transfer with a bit of styling on top is the single most common way these overrun.

What is left is the case Webflow is genuinely good at: a marketing site, a blog, some landing pages, a team that wants to publish without waiting on a developer, and a maintenance burden nobody enjoys. If that is your site, keep reading.

What does not transfer

This is the section to read twice.

Plugins do not come with you. Webflow's own migration page frames this as a benefit, and from a security and maintenance point of view it genuinely is. But every plugin doing real work is a job on your list, which is why step 1 below is an audit of them.

Your theme does not come with you. There is no import path that turns a WordPress theme into a Webflow layout. Anything built with a page builder is markup that gets rebuilt, not content that gets moved.

Files cannot be imported by CSV at all. Webflow's documentation is explicit that no data can be mapped to a file field. If your posts carry PDFs, whitepapers or downloads, those are uploaded by hand afterwards.

Comments have nowhere to go. Webflow has no native commenting. If your comments matter, that is a third party service and a decision about whether the existing thread survives.

Rich text has to be HTML. Webflow will import a rich text field, but only if the value is written as HTML. Custom code embeds inside rich text are supported and then, in Webflow's own words, may be cleaned up to remove elements, classes and attributes it does not support. Anything clever inside your post bodies needs checking after import rather than assuming.

Video fields only accept YouTube and Vimeo URLs. Self-hosted video, or anything on another platform, will not map.

Numbers are fussier than you expect. A number field rejects values containing letters, symbols or commas, so "$100", "100 USD" and "10,000" all fail. If your custom fields hold prices or large figures, they need cleaning before import.

Images imported by CSV get no responsive variants. This one is easy to miss and shows up later as a slow site. When images arrive through a CSV or the API, Webflow does not generate the srcset and sizes variants it would create for an upload. Either optimise and resize before importing, or upload through the Assets panel and accept the manual work.

Custom post types become Collections, and the mapping is yours to design. Nothing decides for you which post type becomes which Collection, or how your taxonomies map onto reference and multi-reference fields. Reference fields match by slug, and Webflow generates that slug from the plain text value, so an item with a custom slug will silently fail to link unless the CSV carries the slug instead.

The migration, step by step

Eight steps, in the order they have to happen:

1. Inventory the plugins and the URLs.
2. Export from WordPress and convert the XML to CSV.
3. Build your Collections and import the content.
4. Rebuild the design rather than porting it.
5. Map every old URL to a new one.
6. Keep the staging site out of search results.
7. Add a Site plan, connect the domain and verify it.
8. Publish, then check the redirects took.

Steps 1 to 6 are reversible. Step 7 is not, which is why it comes so late.

1. Inventory the plugins and the URLs

Two lists, before you open the Webflow Designer.

Go through your active plugins one by one and write down what each does, what replaces it, and what it costs. Most fall into three piles: things Webflow does natively, things a third party service does for a monthly fee, and things that quietly do not get replaced. That third pile is the real decision, and it is much cheaper to find now than in week six.

Then export a full list of every URL the site currently serves. You need it in step 5, and it is also the number that sets your timeline, so knowing whether it runs to 40 rows or 900 changes what you plan for.

2. Export from WordPress and convert to CSV

WordPress exports XML. Webflow imports CSV. There is a conversion step between them, and it is the least interesting part of the whole project.

Read Webflow's CSV import limits before you export rather than after, because they change how you clean the file. Then export your content from WordPress and run the XML through an XML to CSV converter.

Two limits to plan around. The upload caps at 4MB, so a large blog gets split into batches. And if you touch the file in Excel, save it as comma separated values specifically, because the other CSV variants Excel offers do not import cleanly.

3. Build your Collections and import the content

Create a Collection in Webflow, import the CSV into it, and map each column to a field as you go, creating new fields where nothing matches. Webflow's importer tells you whether incoming rows match existing items, which matters more on a second pass than a first.

Check your Collection item limits against your plan before you import rather than after. Premium includes 40 Collections and 20,000 items, more than most sites brought over from WordPress will ever need, but worth knowing rather than assuming.

Do this before you design. It is the order Webflow's own migration guide uses and it is the right way round, because your Collection structure decides what your templates are able to display.

4. Rebuild the design, do not port it

The most expensive mistake in this migration is treating it as a data move with a design task attached.

Your WordPress site's structure reflects what your theme and its page builder made easy. Webflow makes different things easy. Pages that took three stacked plugins to assemble often become one clean layout, and layouts that were trivial in your theme sometimes need rethinking. Teams that try to reproduce the old site pixel for pixel spend the budget fighting the tool and end up with a Webflow site shaped like a WordPress one.

Start from a template or from scratch, decide what the site should be now, and let the old structure inform rather than dictate it. This is also the moment to delete the pages nobody reads, which is free here and awkward later. Bring media in through the Assets panel rather than by CSV wherever page speed matters.

5. Map every old URL to a new one

If your site has search traffic, this is the step that can lose it, and it is entirely preventable.

Take the URL list from step 1, map each row to its new path, and make an actual decision about the pages with no equivalent rather than letting them 404 by default.

Webflow takes a two column CSV of old and new paths under Site settings, Publishing, 301 redirects, up to 16MB. Three things about that import are worth knowing in advance:

1. A new import overwrites every redirect already on the site. If you have existing redirects, export them, add your new rows to that file, and import the combined set. Uploading a fresh file wipes what was there.
2. One invalid row fails the whole import. Webflow emails your file back with guidance on the errors, so this is recoverable, but it is not a step to leave until launch night.
3. Redirects only take effect when you publish. Importing them is not enough, which is why step 8 exists.

6. Keep the staging site out of search results

Every Webflow site comes with a staging subdomain at yoursite.webflow.io, and that is where you get sign-off before anything reaches your real domain. It is public, though, so a half-built copy of your site can be indexed while you are working on it. Webflow has a setting to disable search engine indexing on the staging subdomain. It needs a Site plan or a paid Workspace, so on a free Starter site this waits until step 7. Turn it on the moment you can rather than discovering the duplicate later.

7. Add a Site plan, connect the domain and verify it

You cannot publish to a custom domain without a paid Site plan on the site, so this is where the plan you chose in advance becomes a bill.

If the domain currently points at your WordPress host, clear the old records first: Webflow's advice is to delete the existing A and CNAME records to avoid conflicts you will otherwise spend an afternoon troubleshooting. Then add the domain under Site settings, Publishing, Production, and set the DNS records at your registrar. At the time of writing that is an A record on @ pointing to 198.202.211.1, a CNAME on www pointing to cdn.webflow.com, and a TXT record on _webflow carrying a one time verification value. Copy those from Site settings rather than from any guide, including this one: Webflow has been migrating its DNS infrastructure since April 2025 and the correct records differ between sites.

Then use Check status to confirm the domain resolves, and Verify domain to clear the ownership check. DNS changes can take up to 48 hours to propagate everywhere, though a few hours is more typical, which is the reason this is not a Friday afternoon job.

One more thing here, and it catches people badly. If you have connected more than one domain, set a default so everything else redirects to it, or you have duplicate content. Make the www subdomain the default. Setting the root domain as the default without matching DNS changes produces a redirect loop, and a redirect loop keeps visitors and search crawlers off the site entirely.

8. Publish, then check the redirects took

Publishing pushes to whichever domains you tick in the publish menu, so the staging subdomain and the live domain move independently. Publish to the custom domain once it has verified, then walk your redirect list and check that a sample of the old URLs actually land where you mapped them, because those redirects only became real at this publish.

Keep the WordPress hosting and a database backup running for a while afterwards. It costs one more month and it is the only cheap way back.

What it costs to run afterwards

Webflow's site plans are priced per site in USD, and the important detail for anyone coming from WordPress is that the Basic plan has no CMS. Basic, at $15 a month billed yearly, gives you a custom domain, 300 static pages and unlimited form submissions, and it will not run a blog. Any site with content that updates starts at Premium, $25 a month billed yearly, which is where the CMS, site search and form file uploads live. The free Starter plan is for evaluating, not publishing: two static pages on a webflow.io domain.

Above that the ladder jumps sharply. The Team plan is $2,500 a month on an annual contract, and it exists for organisations that need publishing workflows, localisation, single page publishing and an activity log. There is not much between $25 and $2,500, which is worth knowing if you were expecting a gentle upgrade path.

Compare that against what WordPress actually costs you, not what it appears to. The software is free and GPL licensed, and it powers 41% of the web, but the running total is hosting, plus plugin licences, plus the maintenance labour that usually goes unpriced because it is someone's Friday afternoon. Add the plugin replacements from step 1, since some of what WordPress did for free becomes a subscription on the other side.

For a straightforward marketing site the totals often land closer together than people expect. For a site leaning on a large plugin stack, Webflow is usually more expensive in cash and cheaper in attention, and which of those you are short of is the actual decision.