- View the page source and search for your hero sentence. If it is there, crawlers can read it.
- An empty root div plus a script tag means the page is drawn by JavaScript in the browser.
- Google runs JavaScript in a later step, but a Vercel and MERJ study found many AI crawlers do not.
- Older Lovable apps are pre-rendered for verified crawlers only, so check with Google's URL Inspection tool.
- Check each page has its own title and that a sitemap exists, then fix in place, add server rendering, or rebuild.
The short answer
Open your page, choose View Page Source, and search for your hero sentence. If the sentence is there, crawlers can read it. If you find only an empty div and a script tag, the words are drawn by JavaScript in the browser. Google can run that JavaScript, in a later step, but many AI crawlers do not. Then check that every page has its own title and that a sitemap exists.
One catch for Lovable sites: Lovable says it serves a rendered page to verified crawlers, so your own browser may show the empty shell while Google gets the full page. Use Google's URL Inspection tool to see what Google gets.
Step 1: view source and find your hero sentence
- Open the page in Chrome, right click and choose View Page Source. Or put
view-source:before the URL. - Do not use Inspect. Inspect shows the page after scripts ran, so it always looks fine.
- Press Ctrl+F (Cmd+F on a Mac) and search for a few words from your hero sentence.
- Then search for a price, an FAQ answer and the name of one inner page.
Do this on the home page, a service or product page and one blog post. Different page types are often built in different ways.
What an empty root div means
If view source shows something like this, the page is a client-rendered app:
<body>
<div id="root"></div>
<script type="module" src="/assets/index.js"></script>
</body>The first HTML holds an empty container and a script. The browser downloads the script, runs it, and only then draws your heading, copy and links. People never notice, because it happens in a moment. A crawler that does not run JavaScript sees an empty page with a title.
Google's guide calls this the app shell model: the first HTML does not contain the content, and Google has to run JavaScript to see it. Our glossary has short entries on client-side rendering and server-side rendering.
Lovable, v0 and Bolt: what the docs say
Do not assume a tool always makes one kind of site. Check what yours does. Here is what the vendors' own docs say today:
- Lovable. Lovable's SEO page says apps created from 13 May 2026 (22 June 2026 on Enterprise plans) use TanStack Start with server-side rendering, so every request gets full HTML. Older React and Vite apps get pre-rendering on deployed public URLs, but only for verified crawlers such as Google, Bing and the AI search crawlers. Everyone else, including you in a normal browser, gets the single-page app. Lovable suggests Google's URL Inspection tool to check what crawlers receive, and notes that content that loads late may be missing.
- v0. v0's FAQ says v0 uses Next.js. In the Next.js App Router, pages are Server Components by default, so they render on the server. Parts that fetch their data in the browser after load can still be missing, so run the test anyway.
- Bolt and other tools. The output depends on the framework and setup the project uses. Run the same test.
If you exported your code and host it somewhere else, test again. Lovable describes its pre-rendering for sites it hosts, so it may not come with the code.
How Google handles JavaScript
Google's JavaScript SEO basics describe three phases: crawling, rendering and indexing. Googlebot queues pages for rendering, and it is not obvious when a page is still waiting in that queue. When it renders, it runs the JavaScript and reads the result.
So an app shell is not invisible to Google. It is slower and less certain. Google still says server-side rendering or pre-rendering is a great idea, because it makes the site faster for people and crawlers, and because not all bots can run JavaScript.
Two more points from the same guide matter for AI-built sites:
- Google finds links only in
aelements with anhref. Buttons that change the page with a click handler are not links to Google. - To see what Google saw after rendering, use the URL Inspection tool in Search Console, or the Rich Results Test, and look at the rendered HTML.
How AI crawlers handle JavaScript
The best public data is a study by Vercel and MERJ, published in December 2024. It found that none of the major AI crawlers rendered JavaScript, naming crawlers from OpenAI (OAI-SearchBot, ChatGPT-User, GPTBot), Anthropic (ClaudeBot), Meta, ByteDance and Perplexity. Gemini uses Googlebot's setup and renders JavaScript, and AppleBot renders too.
Bots change, and that study is a snapshot. The safe working rule: if a sentence is not in the first HTML, assume many AI assistants cannot read it, unless your host serves them a rendered copy. Our post Can AI crawlers see your site? has a fuller test, including how to fetch a page as a named bot.
Check 2: one title for every page
AI-built sites often ship the same title and meta description on every page, set once in the main HTML file. Search results then show the same line for your home page, pricing and blog. Google asks for a unique, descriptive title on each page.
- View source on three different pages.
- Search for
<title>andname="description". - Compare. If they match, or if they say the tool's default, fix them.
Lovable's own docs say a new project's starting metadata is generic. If titles are set by a script after load, the source shows the default, and only crawlers that render see the right one. Put each page's title in its first HTML. Our free meta tag tool writes the tags for you.
Check 3: the sitemap
Open yoursite.com/sitemap.xml. You want a list of every page you want found, with your real domain, not a preview address. Lovable says it creates a robots.txt for every new project, but not always a sitemap.
Google says a small site with good internal links may not need a sitemap, but a sitemap still helps search engines find pages. On an app shell, links may only exist after the script runs, so a sitemap is worth having. Our post on what to put in an XML sitemap covers the details.
While you are there, open /robots.txt and make sure it does not block the whole site.
Read your results
| What you find | What it means | Next step |
|---|---|---|
| Hero sentence in view source, own title per page, sitemap live | Readable by search engines and AI crawlers | Work on content and links |
| Empty root div, but URL Inspection shows full HTML | Your host serves crawlers a rendered copy | Fix titles and sitemap, keep checking |
| Empty root div, URL Inspection also thin | Crawlers get little or no content | Turn on server rendering or pre-rendering |
| Same title on every page | Search results cannot tell pages apart | Set a title and description per page |
| No sitemap | Pages are harder to find | Add one and submit it in Search Console |
Next steps: fix, prerender or rebuild
- Fix in place. Titles, descriptions, the sitemap, robots.txt and real
alinks can usually be fixed inside the tool, often with a prompt. - Server render or pre-render. If the words are missing from the HTML, switch the project to server rendering or pre-rendering. On Lovable, that can mean upgrading an older project to TanStack Start, which Lovable's docs describe.
- Rebuild. If the site is mostly marketing pages, a platform that outputs plain HTML is often simpler to run. We rebuild AI-built sites in Webflow or Astro, keeping your URLs and copy: see AI-built site rebuild.
Want a quick read first? Our free AEO checker reads the raw HTML of up to five pages and lists what is missing. Or send us the URL and we will tell you which of the three fits.
Terms used: Sitemap, JavaScript SEO



