Bot vs User View

What crawlers see vs what people see

Enter a URL and we fetch it two ways. The bot view is the raw HTML your server returns, before any JavaScript, which is what crawlers get first. The user view is the same page in a real Chrome browser after the scripts have run. Then we compare them side by side:

• Word count in each, and the difference.
• Title and meta description: do they change after rendering?
• Headings: which H1s and H2s exist in both, and which only appear after JavaScript.
• Links: how many only exist in the rendered page.
• Text differences: samples of content in one version but not the other.

It flags the big risks: no text at all in the HTML, a title that changes after rendering, and headings or links that only exist once JavaScript runs.

Why the gap matters

Search engines work in stages. They fetch the HTML first and can index it straight away. Rendering the JavaScript happens later, if it happens at all. Google renders most pages eventually, but it can take a while, and rendering fails if scripts error, time out, or need a click or a login.

Plenty of systems never render JavaScript. Facebook, LinkedIn, Slack and X preview bots only read the raw HTML. So do many SEO crawlers and AI crawlers collecting content for search and answer engines. For all of them, the bot view is your page.

Reading the results

No text in the initial HTML

The classic client-side app: an empty <div id="root"> and a script tag. The whole page depends on rendering.

Title changes after rendering

Non-rendering crawlers and social apps use the original title; Google may use either. Set the final title in the server response so everyone sees the same one.

Headings that only appear after JavaScript

If your H1 and section headings are injected by scripts, a non-rendering crawler sees a page with no structure. Render them on the server.

Links that only appear after JavaScript

Links are how crawlers find your other pages. Menus, category links and pagination that only exist after rendering can leave whole sections of a site poorly discovered.

A big word count difference

Some gap is normal. Cookie banners, chat widgets, recommendations and third-party reviews often only appear in the user view. What matters is whether your main content, the part you want to rank, is in the bot view.

Closing the gap

• Server-side rendering or static generation (Next.js, Nuxt, SvelteKit, Astro, Remix, Angular SSR) is the reliable fix.
• Pre-rendering services can serve bots a rendered snapshot, but they add cost and complexity. Treat them as a stopgap.
• Fetch SEO-critical data on the server. If product details or article text come from an API call in the browser, move that call.
• Keep metadata out of tag managers. Titles, canonicals and structured data injected by a tag manager aren’t in the bot view.

After a fix, run it again. The SEO inspector adds a screenshot, meta tag checks and performance data to the same comparison.

Quick answers

Which user agent does the bot view use?

It identifies as the SEOLamp crawler. If you think your site serves different HTML to Googlebot, compare with URL Inspection in Google Search Console, which shows exactly what Googlebot received.

Google renders JavaScript, so why worry?

Not instantly, not always successfully, and Google is only one of the things reading your pages. Putting key content in the HTML removes the risk.