Configure page and site search settings, draft metadata with AI, validate structured data, and understand exactly what static and WordPress exports receive.
Where: Project Settings → Pages and Search engines.
Before you start
- A website project with editable page settings.
- A site address before relying on canonical URLs or structured data.
Key ideas
- Site and page controls are layered: The site address provides the canonical origin and the site-wide search toggle can noindex all output. Each page separately controls export inclusion, homepage status, search visibility, slug, metadata, social image, favicon, structured data, and injected code.
- The AI draft remains editable: The SEO assistant reads page text and proposes title, description, and keywords together as one undoable settings change. It does not publish, choose the canonical domain, or prevent manual editing.
- Structured data has safe fallbacks: A blank field can use automatic Organization, WebSite, page, or Article data. AI can draft richer JSON-LD. Invalid custom JSON falls back to the automatic graph instead of emitting broken markup.
- Delivery targets differ: Canonical, Open Graph, and robots metadata are emitted for HTML deliveries. JSON-LD is emitted only for static simple websites; WordPress, apps, and widgets omit it because runtime content could make a build-time graph inaccurate.
Steps
- Open Search engines and set the public site address. Leave it empty only when a connected domain supplies the final origin.
- Choose whether the whole site should be hidden from search engines.
- Open Pages, select a page, set homepage status where available, and choose whether it is included in publish/export.
- Set the URL slug. The homepage exports at the root; resolve reserved or duplicate slugs.
- Enter title, description, keywords, Open Graph image, and favicon, or choose AI SEO after the page has meaningful copy.
- Review and edit the AI draft so it describes delivered content rather than editor-only text.
- Add valid JSON-LD or keep the automatic graph. Set the site address first so entity and page URLs are complete.
- Add head, body, or footer code only when the target delivery supports it and the code is trusted.
- Export and inspect metadata, sitemap inclusion, canonical URL, assets/images paths, and—on static sites—the JSON-LD graph.
Per-page fields
- Set as homepage (except free canvas)
- Include in publish/export
- Hide from search engines
- URL slug
- SEO title
- Description
- Keywords
- Open Graph image
- Favicon
- AI SEO draft
- Structured data JSON-LD
- Head, body, and footer code
Slug and sitemap behavior
- Homepage exports at /.
- An empty non-home slug can follow the page name.
- Reserved and duplicate slugs are rejected.
- Excluded and noindex pages are omitted from the sitemap.
Structured-data output
- Automatic graphs can include Organization and WebSite.
- CMS posts can become Article entries with dates.
- Valid custom or AI JSON-LD enriches automatic data.
- Invalid JSON falls back safely.
- JSON-LD is limited to static simple websites.
Metadata fallback order
- A filled page value wins.
- A blank page value falls back to General → Publish Defaults.
- If no title default exists, Brand → Name is preferred to the internal project name.
- Clearing a page value does not intentionally export an empty title when a valid fallback exists.
Recent export correctness
- Static output excludes the preview-only f0-page container.
- WordPress retains the page container required by that integration.
- Exported images return to assets/images.
- Signed-in redirect behavior is injected at build time.
Tips
- Write real page content before asking AI for metadata.
- Use one public canonical origin and test root and nested URLs.
- Verify that structured-data claims match visible content.
Limitations and important notes
- AI SEO needs enough meaningful page text.
- JSON-LD is intentionally omitted from WordPress, app, and widget output.
- A deployment platform can still override routing or headers.
Troubleshooting
The AI SEO draft is unavailable or weak.
Add meaningful page copy first, then retry.
Custom JSON-LD is absent in WordPress.
This is intentional; configure runtime-aware schema in WordPress.
A page is missing from the sitemap.
Enable export, disable page noindex, and verify the whole site is not noindex.
Images use an unexpected path.
Regenerate with the current exporter; website images now live under assets/images.