Are no-follow links truly ignored by Google when it comes to crawling?

Is no-follow still an absolute blocking signal?

Mueller's statement breaks a firmly held belief for years: the no-follow would completely block the crawl of a URL. The reality is more nuanced. Google may decide to follow a no-follow link if other signals encourage it — for example, if the URL is already known through other sources, or if the context justifies exploratory crawling.

This flexibility is not exhaustively documented. Google uses no-follow as a recommendation, not as an absolute directive. In other words, you suggest to Googlebot not to follow the link, but it may override this based on its own judgment. This behavior has been confirmed by several practitioners who observe crawls on URLs that are exclusively linked with no-follow.

Why do JavaScript on-click links not pass?

Links that use an on-click handler without an <a href="..."> tag are not considered links by Googlebot. Even if the bot renders the JavaScript and executes certain events, it does not detect a link if it does not exist in the DOM as an <a> element with a valid href attribute.

This point is crucial for sites that have switched to modern JavaScript frameworks. If your menus, calls-to-action, or internal links are implemented via div or button with on-click, Google simply does not see them. JavaScript rendering does not compensate for a failing link architecture. This is a common mistake on poorly configured React, Vue, or Angular sites.

What are the implications for internal link architecture?

This statement challenges two widespread practices. First, the use of no-follow to control crawl budget: if Google can still follow the link, you lose a portion of your control leverage. Second, the implementation of links via JavaScript: if you rely on JS rendering for Google to discover your pages, you are taking a risk.

The issue arises especially for sites with thousands of pages. If you no-follow entire categories thinking they won't be crawled, but Google decides to follow them anyway, you end up with a dispersed crawl budget on pages you specifically wanted to avoid. On the JavaScript side, the solution is simple: every important link must exist in the initial HTML, not just after JavaScript execution.

• No-follow is not a guarantee against crawling — Google can follow the link if other signals encourage it.
• On-click links are not recognized as links — they do not appear in Google's link graph.
• JavaScript rendering does not replace a valid href — even if the bot executes the JS, it does not detect a link without an <a> tag.
• Link architecture should be considered in native HTML — JavaScript can enrich the experience, but not replace base links.
• Crawl control via no-follow is less reliable than with robots.txt or noindex — prefer these methods for truly blocking a URL.

Is this statement consistent with what is observed on the ground?

Yes, and that’s what makes it interesting. For years, SEOs have reported crawls on URLs exclusively linked with no-follow. Google had never really clarified this behavior — the official line remained vague. Here, Mueller confirms what the server logs were showing: the no-follow is a hint, not an absolute directive.

The problem is that we still do not know exactly under what circumstances Google decides to follow a no-follow. Is it linked to the popularity of the URL? The depth of the site? A random exploratory crawl? [To be verified] — Google provides no specific criteria. This lack of transparency makes it difficult to base any crawl control strategy solely on no-follow.

Are JavaScript links really invisible in all cases?

Mueller's statement is clear: on-click handlers without href are not links. But it should be nuanced. If a site uses a modern framework with a client-side router and links are dynamically generated with valid <a href="..."> in the DOM after rendering, Google can see them.

The real issue arises when the link exists only as an event listener on a non-semantic element. For example, a <div onclick="goToPage('/product')"> will never be crawled. Even if Googlebot executes the JavaScript, it does not detect this type of navigation. This is often seen on e-commerce sites that have migrated to React or Vue without considering link architecture.

Should one abandon no-follow to control crawling?

No, but it should be combined with other levers. The no-follow remains useful to signal to Google that a link is not editorial — that’s actually its initial role. But if your goal is to strictly block the crawling of a URL, prefer robots.txt or a noindex tag.

Let’s be honest: relying solely on no-follow to manage your crawl budget is risky. Google can still decide to follow the link, and you have no way to predict this behavior. If a page truly should not be crawled, use a Disallow in robots.txt. No-follow remains relevant for outbound links, UGC, or sponsored links — but not as the only internal control mechanism.

What should you do with internal no-follow links?

First step: audit your internal no-follow links. If you are using this tag to control crawl budget, check your server logs to see if Google is still crawling those URLs. In many client sites, we observe that 20 to 30% of no-followed URLs are still crawled anyway. This is a waste of budget if these pages have no SEO value.

If you truly want to block a URL, switch to robots.txt or noindex. The no-follow remains useful to signal non-editorial links — comments, widgets, outbound links — but do not rely on it as a crawl barrier. And if you are no-following links to important pages just to manage depth, rethink your architecture: it’s better to improve internal linking than to no-follow indiscriminately.

How to ensure your JavaScript links are crawlable?

Use the Search Console and URL Inspection Tool to see the DOM rendered by Google. If your links do not appear as valid <a href="..."> in the rendered HTML, they are invisible. You can also analyze the initial source code: every critical link must exist before the execution of JavaScript.

Specifically, if you are using a JS framework, ensure that your router generates <a> tags with href in the DOM. No clickable divs, no navigation driven purely by JavaScript without href. And test with a browser that has JavaScript disabled: if your links stop working, Google probably doesn't see them either. It’s a brutal but effective test.

What mistakes should be avoided in managing internal links?

A classic mistake: massive no-following to 'optimize' internal PageRank. This PageRank Sculpting is outdated for over ten years. Google ignores no-follow links for PageRank calculation, so you gain nothing — and you risk blocking URLs that should be crawled. It’s better to build a logical link architecture than to fiddle with no-follow everywhere.

Another pitfall: implementing menus or calls-to-action solely in JavaScript without a fallback HTML link. This is seen on sites that have migrated to modern frameworks without prior SEO audits. The result: orphan pages, invisible categories, a collapsed internal linking structure. If you are redesigning your site, always keep a layer of native HTML links.

• Audit server logs to identify no-follow URLs that are still being crawled
• Replace no-follow with robots.txt or noindex to truly block a URL
• Check in Search Console that all important links exist as <a href> in the rendered DOM
• Test site navigation with JavaScript disabled to spot invisible links
• Ban clickable divs or buttons without href for critical internal links
• Review link architecture if you are massively no-following to control crawl