Why isn't my Lovable site showing on Google?

First, find out which kind of Lovable app you have

Lovable apps come in two kinds, and the fixes differ. Apps created on or after May 13, 2026 run on TanStack Start, which builds each page on the server. New Enterprise projects followed on June 22, 2026. Older apps are React apps built with Vite. They send an almost empty page, and the browser fills it in with JavaScript.

To check yours, open your live site and view the page source (Ctrl+U, or Cmd+Option+U on a Mac). Or run this in a terminal with your own address.

curl -s https://www.example.com/ | grep -o '<div id="root"></div>'

If the source shows your headings and text, the app renders on the server. If the page body is only <div id="root"></div>, it is an older app. You can also ask Lovable in chat: "What stack is this project on?"

Google usually gets more than the page source shows

Lovable serves pre-rendered copies of published pages to the crawlers it verifies. Its list includes Google, Bing, social preview bots, and AI engines such as ChatGPT, Perplexity, Claude, and Gemini. So on an older app, Googlebot normally gets your text, even though the page source is empty.

SEO checkers and other crawlers get the empty page. That is why an outside audit can call your Lovable site blank while Google indexes it fine. Check what Google itself got before you change anything.

1. In Search Console, paste the page address into the search bar at the top. This opens URL Inspection.
2. If it says URL is on Google, click View crawled page to see the HTML Google indexed.
3. Click Test live URL, then View tested page, and open the Screenshot tab. This is the page as Google renders it today.
4. A blank screenshot or a loading spinner means Google isn't getting your content. See Page indexed without content.

Missing pages that answer 200

On the older Lovable app we tested in October 2026, an address with no page behind it returned status 200, the same as a real page. The app then drew its not-found screen with JavaScript. Google can report addresses like that as Soft 404. The TanStack Start apps we tested returned a real 404.

Lovable's docs don't say which status its pre-rendered copy sends Google for a missing page. If Search Console lists Soft 404 pages, fix it in the app. Google's JavaScript guide gives two ways: send the visitor to an address that returns 404, or add a noindex robots tag to the not-found screen.

When someone opens an address that has no page, show the not-found screen and add <meta name="robots" content="noindex"> to it. Don't add that tag to any other page.

The lasting fix: upgrade to TanStack Start

After the upgrade, every visitor and every crawler gets the finished page, not only the crawlers Lovable verifies. Lovable says link previews get more reliable too.

1. In chat, type / and pick Migrate to TanStack Start. Or open Settings > General > Project actions and select Migrate.
2. Select Start migration on the card that appears.
3. Click through your pages in the preview. Lovable warns that libraries made only for the browser can break server rendering.
4. Publish. Your live site keeps the old version until you do, and you can go back from version history.

Lovable says the upgrade usually costs 10 to 35 credits, depending on the size of the project.

Titles, sitemap, robots.txt, and canonical tags

Lovable's own review covers most of these. Open More > SEO & AI search and click Scan project. Click Try to fix on a finding, and Lovable makes the change in chat. The scan is free, and fixes use message credits. The review doesn't rerun when you publish, so click Scan again after big changes.

Custom domain: give Google one address

Connect your domain in Project > Settings > Domains. The first custom domain becomes the primary domain. Once it shows Live, Lovable sends visitors from your lovable.app address and any other domain to it.

Lovable uses a 302 redirect for this, because you can change the primary domain later. Google doesn't take a 302 as a sign that the new address should be the main one, and it can keep showing the old address. So set a canonical tag on every page that points at your own domain, and submit only your own domain's sitemap.

Lovable's Search Console connector can verify your site and submit the sitemap. It verifies only with the HTML tag, and the tag must be in the HTML the server sends. On an older app, a tag added with JavaScript fails with failedToFindMetaTag. Branded workspace URLs and unpublished projects never get indexed.

What Search Console will show

Three reasons in Indexing > Pages trace back to how Lovable apps work.

1. 1Soft 404

   Why it happens

   Older Lovable apps answer 200 for addresses with no page, then draw a not-found screen with JavaScript.

   The fix

   Add a noindex robots tag to the not-found screen, or send visitors to an address that returns 404. TanStack Start apps returned a real 404 in our test.

2. 2Crawled - currently not indexed

   Why it happens

   Google read the page and chose not to keep it. Short pages built from a single prompt give it little of their own to keep.

   The fix

   Give each page one clear topic and real detail, and link to it from your home page. Ask Lovable to write from your own notes, not generic copy.

3. 3Page with redirect

   Why it happens

   Once a primary domain is set, your lovable.app addresses and other domains redirect to it. Search Console lists those old addresses here.

   The fix

   This is expected. Submit only your primary domain's sitemap, and keep every canonical tag on your own domain.

How CrawlCoach helps on a Lovable site

CrawlCoach reads Search Console's verdict for each page in your sitemap and groups the pages Google skipped by reason. On a Lovable site, that puts the Soft 404 and Crawled - currently not indexed pages in one list, each with its fix. No sitemap yet? The free report says so, and checks the pages Search Console already knows instead.

Paid plans also read the HTML your server sends, without running JavaScript. On an older Lovable app, they flag pages people find on Google whose text isn't in that HTML. Crawlers outside Lovable's verified list see those pages empty.