A hidden HTTP page can cause a naming issue in Google

John Mueller from Google revealed an unusual problem: an old, invisible HTTP homepage can cause issues with the site name and favicon display in Google search results.

Context: a site used HTTPS, but a default HTTP homepage remained accessible on the server. The catch? Chrome automatically upgrades HTTP requests to HTTPS, making that HTTP page invisible during normal browsing. However, Googlebot does not follow that behavior and indexed the wrong version. Google determines the site name and favicon from the homepage by reading structured data, title tags, heading elements and other signals. If Googlebot reads a default HTTP page instead of the actual HTTPS page, it uses the wrong information.

John Mueller recommends two methods to see what Googlebot really sees:

• Use the curl http://yourdomain.com command in the terminal to show the raw HTTP response without Chrome's automatic upgrade
• Use the URL Inspection tool in Search Console with a live test

If the response returns a server default page instead of your real homepage, that’s the problem

Source : Search Engine Journal

This case perfectly illustrates why a technical audit cannot always be limited to automated tools, even for a “all-HTTPS” site.

Goossip #2

Don't serve "not available" via JavaScript

John Mueller strongly advises against showing “not available” via JavaScript before the real content loads. This practice can make Google believe the page doesn’t exist, preventing its indexing and ranking in search results. He instead recommends loading the entire content block directly via JavaScript.

If a client (like Googlebot) does not execute JavaScript or only executes it partially, it will receive misleading information indicating that the content is not available. When Google crawls the page, it sees only the “not available” message and leaves, without waiting for another message to appear.

John Mueller compares this situation to Google’s guidance regarding noindex tags in JavaScript. Google discourages using JavaScript to change a meta robots tag from “noindex” to something else (and there is no “index” tag, only the absence of noindex).

Source : Search Engine Roundtable

From an SEO perspective, John Mueller’s recommendation is common sense. Still, it can be frustrating to see Google continue to penalize sites for JavaScript rendering issues in 2026.

Goossip #3

Use visible anchor text for your links

John Mueller recommends always prioritizing visible anchor text for links to provide more context to search engines. In other words, don’t rely solely on the title attribute in links; make sure the links include actual visible anchor text.

Source : Search Engine Roundtable

This recommendation isn't new, but it serves as a reminder of a fundamental SEO principle: Clarity and visibility above all. It's better to avoid leaving Google to guess a link's intent from hidden attributes (title, aria-label). It's safer to make the visible text explicitly indicate what the destination page is about.