To export your Yoast SEO data when leaving WordPress, you have to know where it lives: Yoast stores your SEO titles, meta descriptions, canonical tags, focus keywords, and schema settings in WordPress custom fields, not in the visible content. That is why they do not show up in an ordinary content export and are so easy to lose. The reliable way to capture them is to crawl the live site, which records the final titles, descriptions, and schema Yoast actually outputs, mapped to each URL. WPBuildAI extracts that rendered metadata for every page and carries it into the rebuild, so your SEO settings move with the content.
One exported field is worth less than it used to be: Google confirmed FAQ rich results stopped appearing on 7 May 2026, with Search Console reporting retired in June and API support ending in August 2026, so FAQ blocks carried out of Yoast no longer need protecting as a rich-result asset.
Where Yoast keeps your settings
Most of Yoast’s per-page data sits in the WordPress postmeta table under fields prefixed with _yoast_wpseo_, the SEO title in one field, the meta description in another, the canonical, the focus keyword, the robots directives, each as its own row keyed to the post. Site-wide settings, your title templates, schema defaults, and social settings, live separately in the options table. This is why Yoast data is invisible to a normal content export: it is not part of the post body, it is metadata attached alongside it. You could query those fields directly if you are comfortable in the database, but the raw fields often contain templates and placeholders (like %%title%% %%sep%% %%sitename%%) rather than the final text, so you would still have to resolve them to know what Yoast actually outputs.
Why crawl the rendered output instead of the database
The simpler and more accurate route is to crawl the rendered pages, because what Yoast writes into the page head is the truth that search engines read. When a page loads, Yoast resolves its templates and fields into the final title tag, meta description, canonical link, robots meta, Open Graph tags, and JSON-LD schema block. Crawling captures exactly that finished output, per URL, with no template placeholders to interpret. So instead of pulling _yoast_wpseo_title and guessing how the template renders it, you read the actual <title> the page produced. This gives you a clean, accurate map of URL to metadata that a new site can reproduce directly, which is precisely the dataset a faithful migration needs. The rendered output is both easier to capture and closer to what matters, what Google actually sees.
What to capture per URL
A complete metadata export records, for each URL: the SEO title, the meta description, the canonical URL, the robots directives (index/noindex, follow/nofollow), the Open Graph and Twitter card tags, and the JSON-LD schema block. Each of these does a job. The title and description shape how the page appears and is understood. The canonical prevents duplicate-content confusion. The robots directives control whether the page is indexed at all, miss a noindex and you might accidentally index pages you meant to keep out, or vice versa. The Open Graph tags control social sharing previews. And the schema tells engines what the page is. Capturing all of these per URL, rather than just the title and description, is what makes the export complete enough to rebuild the page’s full SEO footprint, not just its headline.
Why this metadata matters so much
It is small in size and large in effect. Backlinko’s analysis of 11.8 million search results found title tags and clear on-page signals correlate with ranking, and the 2024 Web Almanac shows how widely structured data is now deployed, which means losing your schema and titles in a move puts you behind the norm rather than at it. Google’s introduction to structured data explains how that schema can drive rich results, which you forfeit if it does not carry across. A page that loses its title and description looks thin and ambiguous to Google overnight, which is a frequent, quiet cause of post-migration drops, the failure traced in why traffic drops after a redesign. The metadata is invisible to visitors and decisive to search engines.
A worked example: 300 pages of metadata
Picture a 300-page site leaving WordPress. The owner exports the content as Markdown and rebuilds it, and the pages look right, but two weeks later rankings slip. The cause: the rebuild carried the body text but not the Yoast metadata, so every page lost its crafted title and description and reverted to a generic auto-title, and the article schema vanished. A crawl-based metadata export would have prevented this: it captures all 300 pages’ titles, descriptions, canonicals, robots directives, and schema as a URL-keyed table, which the new site applies page by page. Done that way, the 300 pages keep the exact titles and descriptions they ranked with, and the schema continues to drive rich results. The body content was never the risk; the invisible metadata layer was, and exporting it deliberately is what kept the rankings intact through the move.
Carry it into the new site
Exporting the metadata is half the job; applying it on the new site is the other half. Reapply the captured title, meta description, canonical, robots directive, and schema to each corresponding page on the new platform, and keep the URLs the same or 301 them, per Google’s site move guidance. On an AI-built or modern site this matters doubly, because many builders ship without a metadata layer at all, so the export is the only way the new pages get proper titles and schema, the issue covered in an AI builder site missing schema and meta tags. The metadata belongs on the migration checklist alongside URLs and tracking tags, as in the WordPress migration SEO checklist and next to the kind of break behind stopped analytics. Export and reapply are one task in two halves.
Do not forget site-wide and template settings
Per-page metadata is the bulk of the export, but Yoast also holds site-wide and template-level settings that shape pages you did not configure individually. The title templates that generate titles for posts, archives, and taxonomies; the default schema for the organisation or site; the social defaults; and the indexing rules for categories, tags, and author archives all live in Yoast’s options rather than per page. If you only export per-page metadata, the pages that relied on templates, often archives and taxonomy pages, end up with no metadata on the new site. So capture the patterns too: how titles were generated, which page types were noindexed, what the default schema was. The crawl helps here, because it shows the rendered result of those templates on real pages, from which you can infer and reproduce the rules.
Common mistakes exporting Yoast data
The recurring errors all come from underestimating where the SEO data lives. Relying on the WordPress XML export misses the Yoast fields entirely, since they are metadata, not content. Pulling the raw _yoast_wpseo_ fields without resolving their templates leaves you with placeholders instead of final text. Capturing only titles and descriptions drops the canonicals, robots directives, and schema that also matter. Forgetting the site-wide templates leaves archive and taxonomy pages bare on the new site. And exporting the metadata but never reapplying it on the rebuild, treating the export as the finish line, loses it anyway. Each is avoided by crawling the rendered output for the complete per-page metadata, capturing the template patterns too, and reapplying everything on the new site with URLs preserved or redirected.
Key points to remember
Yoast stores your SEO titles, descriptions, canonicals, robots directives, and schema in WordPress custom fields and site-wide options, not in the visible content, which is why they vanish from an ordinary export and quietly cost rankings when a migration forgets them. Capture them by crawling the rendered pages, which gives the final output per URL rather than raw templates, and record the full set per page plus the site-wide title and schema patterns. Then reapply everything on the new site with URLs preserved or 301’d, because the export is only useful once the new pages carry it. WPBuildAI extracts the rendered metadata for every page and rebuilds with it in place, so your SEO settings survive the move off WordPress; send your site URL for a fixed quote.
Not affiliated with Yoast, WordPress, or Lovable.