The pitch most Webflow migrations get is one paragraph long. "Webflow's slowing us down, the CMS limits are biting, let's move to a custom Next.js build." Six weeks later the new site ships, and within a month organic traffic is down 30 to 40 percent. The agency shrugs and blames Google.

Google didn't do this. The migration did.

I want to be direct about my position first. Moving off Webflow to a custom-coded site is, in most cases, the right call for an Australian business that has outgrown the platform. The performance ceiling is real, the CMS pricing tier jumps are real, and the long-term cost math favours owning the code. But if you migrate badly you'll spend the next eight months recovering the rankings you already had — and that's an expensive own-goal nobody factored into the quote.

This is how I'd run the migration if I were doing it for you tomorrow.

Why most Webflow-to-custom migrations bleed traffic

Ahrefs data from 2024 found that sites which migrate URLs without proper 301 redirects lose on average 38 percent of organic traffic within 30 days. That's the headline number every Australian business owner should sit with for a second before signing a migration quote. Nearly two-fifths of your Google traffic, gone in a month.

The usual failure modes:

The old URL structure isn't fully mapped. Maybe the homepage and the top ten pages are covered, but the long tail of blog posts, service pages, and product detail URLs gets dropped.
The redirects are implemented but not 301s. Webflow's redirect panel and Next.js's next.config.js both handle this fine, but agencies sometimes ship 302s (temporary) by accident, which don't pass link equity.
Metadata gets rebuilt from scratch. Title tags and meta descriptions that took years to optimise get replaced with the developer's first draft, because the export didn't carry them across.
Internal links still point at the old URLs. The redirect chain works, but every internal click costs a redirect hop, which both wastes crawl budget and slightly diminishes link equity.
The XML sitemap doesn't get resubmitted in Google Search Console on launch day, so Google takes weeks longer to discover the new structure.

None of these are clever traps. They're basic technical hygiene. They get missed because the migration timeline is too tight and nobody is specifically responsible for SEO continuity.

The case for moving (and being honest about it)

Before the playbook, the nuance. Webflow is a legitimately good platform for a chunk of small business websites. The visual editor lets a non-developer push content updates. The CMS is fine for a few content types. The hosting is fast enough for most marketing sites.

The reasons businesses outgrow it are specific:

CMS pricing escalates fast. The Business plan ($49 USD/month) caps you at 10,000 CMS items and 50 reference fields, which sounds like plenty until you're a directory site or have a complex services taxonomy. Enterprise pricing is "contact us" territory.
Logic and interactions hit a ceiling. Anything beyond the built-in interactions library — multi-step forms with conditional logic, custom checkout flows, gated content with proper auth — gets bolted on with custom code that fights the platform.
Performance is decent, not great. A well-built Webflow site can hit Lighthouse 80s. A well-built Next.js 16 site with Cache Components sits in the 95+ range without trying. The gap matters most for sites where Core Web Vitals affect rankings.
Ownership is partial. You own your content. You don't own the rendered HTML or the hosting infrastructure. If Webflow doubles its prices tomorrow, your options are pay up or migrate.

If two or more of those bullets describe your situation, moving makes sense. Just don't move badly.

The pre-migration audit (week 0)

Before you touch a line of code, you do the audit. Skipping this is where the 38 percent traffic loss starts.

Export every URL

Pull every URL that's ever ranked. The fastest path: 16 months of Google Search Console performance data exported to CSV, deduplicated, sorted by total clicks descending. That's your priority list — the top 20 percent of URLs drive 80 percent of the organic traffic.

Now overlay your sitemap. Webflow's sitemap.xml lives at /sitemap.xml and lists every published page. You want union of (URLs that ranked) and (URLs that exist) — anything in either set needs a destination on the new site.

For a typical Australian small business site you're looking at 40 to 200 URLs. For a content-heavy site it can be 500 to 5,000. Either way: spreadsheet, column A is the old URL, column B is the new URL, column C is the redirect status, column D is who's responsible for verifying it.

Capture metadata

For every URL in the spreadsheet, pull the existing title tag and meta description. Screaming Frog does this in five minutes for a site under 500 URLs (the free tier covers it). For larger sites, Sitebulb or Ahrefs Site Audit will do the job.

These titles and descriptions are years of accumulated SEO work. You're not throwing them out. The new site inherits them by default and only changes them where you've consciously decided to optimise something.

Map content depth

Webflow's structured content (collections, reference fields, multi-image fields) doesn't always export cleanly. Before the new build starts, document the content model: what types of content exist, what fields each type has, how items reference each other. Your new CMS — whether that's Sanity, Payload, or Contentful — needs to model this faithfully or content editors will revolt the first week.

The build phase (weeks 1 to 6)

The new site gets built on a staging URL nobody crawls. That's important — accidentally indexing a staging environment with the old domain's content creates duplicate-content headaches you don't want.

A few things to bake in from day one:

Match URL structure where possible

If the old URLs were /services/web-design/ and /services/seo/, the new URLs should also be /services/web-design/ and /services/seo/. Every change you make to URL structure is a redirect you'll need to maintain. Make changes only where the old structure was actively bad (deeply nested, query-string-based, missing keywords).

Build the redirect map into code

In Next.js this lives in next.config.js as an exported redirects() function returning an array of objects with source, destination, and permanent: true. The permanent: true flag is what makes it a 301. Get this wrong (302 by accident) and you forfeit the link equity.

For sites with hundreds of redirects, the redirect map becomes a generated file imported into the config. For sites with thousands, you'll want middleware doing pattern-based redirects with a fallback to specific overrides.

Preserve metadata at the component level

Use Next.js's Metadata API to set per-page title and description from your CMS data. The CMS should have a seo field group on every content type with metaTitle and metaDescription. Migrate the values from the audit spreadsheet.

Internal links should point at the new URLs

Not the old ones via the redirect. Update every link in content and navigation to point directly at the new structure. The redirects are a safety net for external links, not the primary navigation path.

The launch (week 7)

Launch day is mechanical. The order matters.

Pre-launch. Final crawl of the staging site to confirm every URL in the redirect map resolves to a 200 response and every old URL 301s to the right new URL.
DNS cutover. Point the production domain at the new hosting. With Vercel this is a TXT record verification followed by an A/AAAA or CNAME change. Propagation is typically 5 to 60 minutes.
Submit the new sitemap. In Google Search Console, submit /sitemap.xml from the new site immediately. Don't wait.
Submit the URL inspection tool for top 20 URLs. The "Request Indexing" feature in GSC for each priority URL kicks the recrawl. It's tedious but it shaves days off recovery.
Set up monitoring. Daily Search Console checks for crawl errors. Daily traffic dashboard in GA4. Weekly ranking checks for top 30 keywords. A noticeable dip in week 1 is normal. A continued slide into week 3 means something's broken.

What recovery actually looks like

Even a well-executed migration sees movement. Search Engine Journal's consensus number — backed by enough case studies to be roughly believable — is that a clean migration sees a 5 to 15 percent traffic dip for two to four weeks, then full recovery by week 6 to 8. A messy migration sees 30 to 50 percent dip and partial recovery only after six months of remedial SEO work.

The difference between those outcomes is the audit and the redirect map. Almost everything else — the framework choice, the CMS choice, the design decisions — matters less for SEO than the boring spreadsheet work that determines whether every old URL has a new home.

The numbers a Webflow-to-custom quote should include

When you're getting quoted for a Webflow migration, the line items that should be in the proposal:

URL audit and redirect map. Should be its own line item, not bundled into "migration." Typically 8 to 20 hours depending on site size.
Metadata preservation. Either a manual export of titles/descriptions or a scripted scrape of the live site, fed into the new CMS.
301 redirect implementation and testing. Including a post-launch verification step.
Search Console handover. Old and new properties verified, sitemap submitted, top URLs requested for re-indexing.
Two weeks of post-launch SEO monitoring. With remedial work included if anything spikes.

If a migration quote doesn't have these as explicit line items, the agency is either bundling them silently or skipping them. Ask which.

The one thing that's not in any playbook

The honest truth about SEO migrations is that some of the recovery is just patience. Google's crawl-and-rerank cycle takes weeks. Even if every redirect is perfect and every meta tag is preserved, ranking signals get redistributed across the new URLs and the algorithm needs to re-confirm authority. There's no shortcut.

What you can control is whether the technical foundations are right so that when Google finishes reshuffling, you land back at the same rankings instead of worse ones. The audit, the redirect map, and the metadata preservation are the levers that determine the outcome.

If you want a baseline on the Webflow site before you move it, run a free audit on the current URL. You'll get the rendered metadata, the mobile and desktop performance, and the specific issues that'll affect ranking continuity post-migration — the same numbers a clean migration plan should be measured against on day one of the new build. Worth doing before you sign the quote, not after.

How to Migrate Webflow to Custom Without Tanking SEO (2026)