Crawled - currently not indexed
Lovable, v0, and Bolt pages that Google never sees
A polished client render can still ship an empty document to Googlebot.
Tools that generate a single-page app often put the marketing copy in JavaScript. The first HTML response is a root node, a loading state, or a few words. Google can render JavaScript, and it often does not wait long enough, or does not index a page whose HTML has nothing to quote.
AI answer crawlers are less patient than Google. If GPTBot or OAI-SearchBot only see the shell, they have nothing to cite.
Move the title, description, H1, and the main paragraphs into the server HTML. A static export, a server component, or prerender of the public pages is enough. Leave the app shell for the signed-in product.
What to check
- Disable JavaScript and read the page. That is close to what a cautious crawler sees.
- Count the words in the HTML. A hero with a handful of words is a thin document.
- Set the canonical and robots meta in that same HTML, not in a client effect.
Fix prompt
Example prompt
This page looks complete in the browser and empty in the server HTML. Render the title, meta description, H1, and main copy on the server for public URLs. Keep client rendering for the signed-in app. Do not call an indexing API.
Check a URL · Crawlers · Glossary