Lovable SEO
Lovable’s SEO problems now depend on when the project was made: newer apps are server-rendered, older React + Vite apps are pre-rendered only for verified crawlers, and in both cases the starting metadata is generic and a sitemap isn’t guaranteed. Find out which stack your project is on, then scan the live URL to see what’s missing.
How Lovable serves pages to crawlers today
Lovable’s docs split projects in two. Apps created from 13 May 2026 (22 June 2026 on Enterprise plans) use TanStack Start with server-side rendering, so each page arrives with its text. Older apps are React + Vite single-page apps, and Lovable pre-renders them on request for verified crawlers: search engines, AI search crawlers, social-preview bots and verified SEO tools. Unverified automated agents get the regular single-page app.
That has a consequence for any scanner, Fecto included. Fecto’s scan identifies itself as FectoScan and isn’t on anyone’s verified list, so on an older Lovable app it sees the empty shell a browser starts with. A failing main-text check there is accurate for unverified fetchers but overstates the problem for Google and the large AI search crawlers. Check the real view with Search Console’s URL inspection before rebuilding anything.
The pre-rendering is part of Lovable’s hosting. If you export an older project and host it yourself, what crawlers get is whatever your new host serves.
Lovable says it creates a robots.txt with every new project, but not always a sitemap, and that the starting metadata is generic. Its SEO & AI search review flags these and can often fix them with Try to fix.
Common causes, and the fix for each
Main text missing from the HTML
Fecto check read.main-text-needs-js · Main text is in the HTML
On a TanStack Start project, a failure here is real: some block loads its text in the browser. On an older React + Vite project, read the caveat above first. Text in the HTML for every fetcher means the server-rendered stack that new projects get. Ask Lovable whether an existing project can move before planning around it; for most sites, verified crawlers seeing the page is what matters.
No sitemap
Fecto check index.sitemap · A sitemap exists
Since a sitemap isn’t guaranteed, check /sitemap.xml on your published domain. Lovable can add one when you ask; the scan confirms it is served and valid.
The generic starting title and description
Fecto check index.meta-tags · Title and description
Every page keeps the metadata Lovable started with until someone changes it. Each page needs its own title, a description written from what it says, and a canonical on your primary domain rather than a lovable.app address.
A share preview that shows the default image
Fecto check index.open-graph · Sharing tags
Shared links show whatever og:image the project started with. Set a real image and text per page, then paste the link into a chat to see it.
Nothing that says who the business is
Fecto check record.no-business-schema · Business described in structured data
Lovable recommends adding structured data where it fits. An Organization block with your name, URL and contact details is the one that fits every business site. Give the details yourself; don’t let the generator fill them in.
Let your agent do it
If the project is synced to GitHub, your coding agent can work in that repo and every fix is a pull request; once you merge it, Lovable picks it up. If you only use the Lovable editor, the agent tells you what to ask Lovable for, check by check, and re-scans the published site after each change.
After changing a custom domain, Lovable says to verify the new domain in Search Console again and resubmit the sitemap. Your agent reminds you; it can’t do that part for you.
No coding agent yet? Install the Claude app, sign in, then come back and choose Claude app.
Set it up by hand
curl -fsSL https://fecto.ai/install | shThen tell it
Use Fecto to check and fix my site.Don’t add a prerendering service on top of Lovable hosting
Lovable says hosted projects don’t need an external pre-rendering service for search engines or AI crawlers. Paying for one adds a second layer that can serve stale copies. And llms.txt isn’t required: Lovable keeps serving one if you have it, and its review doesn’t flag a missing one.
What it costs
You pay only when the result shows up on your site: +20% more visits or enquiries is $30, +50% is $60, double is $100, and anything below +20% costs $0. Never more than $100 a site a month.
Questions
What is a good Lovable SEO prompt?
One per check, naming the result: “add a sitemap.xml listing every public route and declare it in robots.txt”, or “give each page its own title, meta description and canonical on my custom domain”. Then publish and re-scan. A prompt is done when the check passes on the live site, not when Lovable says so.
Where do I see how Lovable pages look in search?
Lovable’s guide points to its Site metadata page for how a page looks when shared and in search, and to the SEO & AI search review for what is missing. The fixes are changes to the project, which is why a prompt or a pull request makes them.
Related guides
Checked against Lovable docs: optimize your app for SEO and AI search on 18 November 2026.