The logo displays normally, yet search results still show a gray globe icon. The problem is usually not whether the image was uploaded, but whether Googlebot has actually crawled, parsed, and cached the icon. Four conditions must all hold at the same time, and the October 27 documentation update turned two of them into hard requirements.
Start with the conclusion: a logo displaying normally does not mean the favicon can enter search results
The logo on the page and the small icon to the left of the site name in search results (favicon) are two independent mechanisms. The former is determined by the template and stylesheet; as long as the image is uploaded successfully and the layout is correct, visitors can see it. The latter can be displayed only after Googlebot crawls, parses, and caches it. Therefore, 'the logo is clearly fine' and 'the favicon does not appear' can absolutely be true at the same time.
The most common misjudgment is treating a horizontal brand image as a favicon. Google does not read design files; it only recognizes the <link rel="icon"> declaration in the homepage HTML and the file it actually crawls. For the favicon to enter the display process, four conditions must all be met: the homepage can be crawled by Googlebot; Googlebot can follow homepage redirects and find the page that declares the favicon; Googlebot-Image can crawl the favicon file; and the file itself meets the official requirements for square shape, size, and format. If any link is broken, the results page will fall back to the default gray globe icon.
What is the difference between a website logo and a favicon?
Confusing the two is the first mistake in troubleshooting. Their purposes, aspect ratio requirements, and who controls them are all different.
| Comparison item | Page logo | Favicon |
|---|---|---|
| Display location | Header, footer, hero section, and other layout positions | Browser tabs, bookmarks, and to the left of the site name in search results |
| Aspect ratio requirement | Can be horizontal, irregular, or even text-only | Requires a 1:1 square canvas |
| Common files | Large PNG, SVG, WebP | Small square images such as ICO and PNG |
| Who determines it | Template structure and stylesheet | Google crops and displays it after crawling |
| Error symptoms | Layout misalignment, image 404 | Default icon shown in results |
Although the icon is small, it carries the second confirmation of brand identity: across multiple tabs, in bookmark lists, and within a screen of search results, it determines whether users can recognize you at a glance. An analysis of the relationship between favicons, branding, and SEO points out that it affects tab recognizability, bookmark discoverability, and users' judgment of a site's professionalism, but it will not significantly improve rankings on its own (DEV Community analysis). A reasonable expectation is that the favicon improves recognition and click intention, rather than doing ranking work for you.
How does Googlebot find a website favicon? Three necessary paths
Understanding the paths is more useful than memorizing scattered tips. Google's order for finding a favicon is roughly three steps:
-
Crawl the homepage: Googlebot requests the homepage (or the canonical homepage URL) and reads the
<head>of the returned HTML. - Follow redirects: If the homepage redirects, Googlebot needs to follow it all the way to the final page and find the favicon declaration there.
- Crawl the favicon file: Googlebot-Image separately requests the favicon URL and confirms that it can be publicly accessed and returns a normal status code.
This also explains the common confusion: the favicon is clearly visible when you open the tab locally. Your browser takes another path; it does not care about robots.txt, nor does it care whether the redirect chain is complete.
The October 27 documentation update: redirects and robots.txt became hard requirements
Google's website favicon documentation updated a key statement on October 27: Googlebot must be able to crawl the homepage, follow any redirects from the homepage to the page declaring the favicon, and Googlebot-Image must be able to crawl the favicon file. These URLs must not be blocked by robots.txt or other means (Search Engine Roundtable report; see the original Google website favicon documentation). Before the change, the wording only required that the homepage and favicon file not be blocked from crawling and did not call out redirects. According to the report, the clarification is related to the fact that the redirect chain is easy to miss when troubleshooting this issue. The implication for business websites is direct: redirects are no longer 'possibly relevant'; they are a prerequisite that must be checked hop by hop.
Where should a favicon be placed? File location and HTML declaration
Putting favicon.ico in the site root is the most compatible fallback; most browsers and crawlers look there by default (DEV Community). However, to precisely control display, it is still recommended to declare it explicitly in the homepage HTML:
<link rel="icon" href="/favicon.ico" sizes="48x48">
<link rel="icon" href="/favicon-96x96.png" type="image/png" sizes="96x96">
<link rel="icon" href="/favicon.svg" type="image/svg+xml">
<link rel="apple-touch-icon" href="/apple-touch-icon.png">
Three details determine whether this code works: the tags must be output by the server; tags injected afterward by JavaScript are likely to be invisible in the first crawl round. One set is enough; when an old theme, SEO plugin, and site icon setting are active at the same time, two conflicting sets of declarations can easily appear. The href must be directly crawlable; a relative path or an absolute CDN address is fine, provided that when Googlebot-Image requests it, it returns the icon itself rather than a 404, a login page, or a redirect back to the homepage. When using platforms such as WordPress, Shopify, or Wix, prefer the platform's built-in site icon setting, then return to the source code to confirm that only one set of tags ultimately lands.
Size and format: why 16×16 is often not enough
Although the icon is rendered extremely small in search results, Google needs a sufficiently large source file to crop. According to a troubleshooting guide organized based on Google's current specifications, Google focuses on square (1:1), a minimum of 8×8 pixels, and recommends larger than 48×48 pixels. Supported formats include BMP, GIF, ICO, PNG, JPEG, PPM, and TIFF, and accepted rel values include icon, shortcut icon, apple-touch-icon, and apple-touch-icon-precomposed (complete fix guide).
In practice, you can make trade-offs like this: place the horizontal logo into a square canvas, or simply use a graphic symbol or initial letter; provide at least 48×48, with 96×96 or 192×192 being more stable; use PNG or ICO for the main file; remove thin lines, small text, and gradient details. SVG works well in browsers but is not in the supported list above. A reasonable approach is to keep SVG for browsers while also providing a PNG or ICO for search engines. In addition, an .ico file may contain only one very small size internally; when troubleshooting, confirm that it really contains a sufficiently large image.
Homepage redirects and robots.txt: two most easily overlooked blockers
Homepage redirects: example.com redirecting to www.example.com, / redirecting to /zh/, or HTTP redirecting to HTTPS are all imperceptible in a browser, but for crawlers they are a path that must be fully traversable. Multilingual sites are especially prone to problems here: an overly long redirect chain, a redirect to a page that requires login, or no favicon declaration at all in the <head> of the final page will break the crawl.
robots.txt: Favicons are often placed in directories such as wp-content, assets, or static, and these directories are the most likely to be blocked by a 'save bandwidth' rule:
User-agent: *
Disallow: /assets/icons/
User-agent: Googlebot-Image
Disallow: /
According to Google's crawler documentation, the user-agent tokens for Googlebot-Image in robots.txt are Googlebot-Image or Googlebot, and any matching rule takes effect. Crawl preferences for Googlebot-Image affect Google Images, Discover, video modes, and all features in Google Search that display images, logos, and website icons (Google crawler list). In other words, a rule written to 'protect images' may very well turn off your search result favicon. There are two ways to fix it: delete the rule, or add a separate Allow: rule for the favicon path.
After changing the favicon, why don't search results update immediately?
Favicons are not read in real time. Google needs to recrawl the homepage, recrawl the favicon file, and then regenerate the display asset. According to the experience in the guide above, a new favicon usually takes days to weeks to appear in search results (ConvertICO). During the waiting period, three things are worth doing: request recrawling of the homepage in Search Console's URL Inspection; keep the favicon URL stable, and do not change the filename on every deployment just to 'force a refresh'; change only one variable at a time, otherwise you cannot tell which step worked. At the same time, accept a premise: even if all rules are met, Google does not promise to display a favicon, and it reserves the right to crop, scale, and replace it.
Website favicon checklist
Compress the sections above into a table that can be checked item by item. When troubleshooting, go from top to bottom and do not skip steps.
| # | Check item | Pass criteria | How to verify |
|---|---|---|---|
| 1 | icon tag in HTML | At least one rel="icon" in the homepage <head>, output by the server |
View page source and search for rel="icon"
|
| 2 | Homepage redirects | The final redirect destination page contains the favicon declaration, and the chain length is manageable | Open the redirects hop by hop and check the status code of each hop |
| 3 | Favicon file accessible | Opening the favicon URL directly returns 200 and image content | Open the favicon address in a new tab |
| 4 | Crawl permission | robots.txt does not block the favicon directory or Googlebot-Image | Check the relevant rules in robots.txt |
| 5 | Size and shape | 1:1 square, at least 48×48, provide 96×96 or larger | Check the image's actual pixels |
| 6 | Format | Primarily PNG or ICO, with SVG only as a supplement | Check the declared type and file extension |
| 7 | URL stability | Filename fixed and accessible over the long term | Compare source code after multiple deployments |
| 8 | Declaration uniqueness | Only one set of icon tags | Search all rel="icon" in the source code |
| 9 | Primary domain coverage | One favicon per hostname; do not expect a separate favicon for subdirectories | Check the www and bare-domain homepages separately |
Treat the favicon as part of the release asset during the site-building stage
The favicon problem is often not a technical challenge but a process problem: the brand logo is repeatedly polished in the design phase, while the favicon is often a file exported temporarily ten minutes before launch, and no one is responsible for verifying how it appears in search results.
A more stable approach is to put brand assets and the favicon into the same publishing workflow: use the same visual source for the favicon and the logo, crop within a square safe area, export multiple versions at 48×48, 96×96, 192×192, and ICO at once, and declare them clearly in the homepage HTML. We0's site-building path is suitable for carrying this out—starting from describing requirements in natural language, a multi-agent team turns the description into a runnable site, fine-tunes details in a visual canvas, and then deploys it to your own domain with one click (We0 AI website builder); the same workflow also covers SEO and GEO optimization, a CMS backend, and multiple design styles. The favicon, site structure, and citable brand information can be confirmed together in one release, rather than waiting until a gray globe icon appears in search results to go back and fix it.
Frequently asked questions
The logo is already set on the page. Why is the default icon still on the left side of search results?
Because the two are determined by different paths. The search result icon depends on the rel="icon" declaration in the homepage HTML and whether Googlebot-Image can crawl that file; simply placing an image in the layout usually does not generate a favicon.
Must the favicon be an .ico file?
No. ICO is only the most compatible fallback format; PNG is perfectly fine and makes it easier to produce high-definition versions. It is recommended to provide both ICO and a PNG of at least 96×96, and leave SVG for browsers.
Will a 301 redirect on the homepage cause the favicon not to be crawled?
The redirect itself is not a violation. The key is whether Googlebot can follow it all the way to the final page and read the favicon declaration there. An overly long redirect chain, a redirect to a page that requires login, or a final page without a favicon declaration will all interrupt crawling.
After changing to a new favicon, how long will it take to see it in search results?
It depends on the time Google needs to recrawl and regenerate the display asset; it usually takes days to weeks. During this period, you can use Search Console to request recrawling of the homepage and keep the favicon URL stable.
Related Tools
- Google Search Console URL Inspection (request recrawling of the homepage)
- The Googlebot-Image entry in Google's crawler documentation (confirm the scope of crawl preferences)
- The Network panel in browser developer tools (confirm the status code and content type of the favicon request)
- An image property viewer (check the actual pixel dimensions and file format)
Related Links
- Google Search Central: Define a website favicon for search results
- Google: Google crawler list (including Googlebot-Image)
- Search Engine Roundtable: Google updates documentation to require Googlebot to follow redirects from the homepage to the favicon
- ConvertICO: Why Is My Favicon Not Showing in Google?
- DEV Community: Why Favicons Matter for Branding & SEO
- We0: AI Website Builder
Summary
The logo displays normally, yet search results still show a gray globe icon. The problem is almost never whether the image was uploaded; it is whether the favicon has been fully crawled: the homepage must be crawlable, redirects must be followable, the favicon file must be accessible to Googlebot-Image, and the file itself must be a compliant square image. All four conditions are required. The October 27 documentation update explicitly made redirects and robots.txt hard requirements. When troubleshooting, check the redirect chain hop by hop and confirm that no rule inadvertently blocks the icon directory. After a fix, you still need to wait for the recrawl cycle and keep the favicon URL stable. Handling the favicon and brand logo in the same publishing workflow is much easier than troubleshooting afterward.
Originally published on We0. This republication was prepared with AI assistance.
Top comments (0)