Noindex, nofollow and robots.txt noindex: what kills a link
A "noindex, nofollow" link in code means one of three different instructions: the rel="nofollow" attribute on a single link asks search engines not to pass weight through it, the <meta name="robots" content="noindex, nofollow"> tag removes the page itself from the index and asks crawlers to ignore all of its links, and robots.txt blocks crawling of the page. They are constantly confused, and each one kills a link in its own way.
What noindex and nofollow look like in code
There is no rel="noindex" attribute for links: a link only has rel with values such as nofollow, ugc and sponsored, while noindex applies to a whole page. So a "noindex nofollow link" in practice means one of the following.
Robots meta tag for the whole page
<meta name="robots" content="noindex, nofollow">It goes in the <head>. noindex keeps the page out of search results; nofollow asks crawlers not to follow the links on that page. You can use either on its own, and you can target one crawler, for example name="googlebot".
Attribute on a single link
<a href="https://example.com/" rel="nofollow">anchor</a>
<a href="https://example.com/" rel="sponsored">paid link</a>
<a href="https://example.com/" rel="ugc nofollow">link from a comment</a>It affects only that link; the page itself is indexed as usual. Separate several values with a space.
X-Robots-Tag HTTP header
X-Robots-Tag: noindex, nofollowThe same rules as the meta tag, sent in the server response. Use it for PDFs, images and other files that cannot carry a meta tag. You can see it with the HTTP headers checker.
You may also meet a <noindex> tag in Russian-language sites: it is a non-standard Yandex tag for hiding a fragment of text from Yandex indexing. Google does not support it, and it is not a substitute for rel="nofollow" on links.
Differences
| Instruction | Where it is | What it does | What happens to your link |
|---|---|---|---|
rel="nofollow" | In the link itself | Requests not to consider the link as a recommendation | There is a mention, weight is not transferred |
noindex | In the page's meta tag or response header | Removes the page from the index | The link is useless: the donor page is not in search |
Disallow in robots.txt | In the file on the domain | Prohibits crawling the page | The robot will not see either the page or the link on it |
rel="ugc" / rel="sponsored" | In the link | Marks the user-generated or advertising nature | Considered as a hint, weight is usually not transferred |
rel="nofollow", "ugc" and "sponsored": nofollow has been a hint since 2019
On 10 September 2019 Google announced on the Search Central blog two new values, ugc and sponsored, and that all three values are now hints for ranking rather than strict directives. For crawling and indexing, nofollow became a hint on 1 March 2020. Hence the caveat in the table above: weight "usually" does not pass through a nofollow link — Google may count it, but you cannot rely on that.
| Value | When to use it |
|---|---|
sponsored | Ads, sponsorships and any paid placement |
ugc | Links in user-generated content: comments, forums, reviews |
nofollow | Other cases where you do not want to vouch for a site or pass it weight |
Paid links that pass ranking credit violate Google's spam policies, which is why they are marked sponsored or nofollow. When you evaluate someone else's platform, a sponsored or ugc link is the same as nofollow: a mention with no guaranteed weight. More in link weight and PageRank.
The worst combination: dofollow on a noindex page
This exact case occurs on past services: the link is honest, without nofollow, but the page itself is closed from indexing. In our pool, there are such platforms, and their share of publications in the index is measured at zero — hence they receive low priority or are turned off altogether. How the selection works is described in the article “How we select platforms”.
rel attribute in the source code often lies: one platform removes nofollow with a script already in the browser, another, on the contrary, adds it. You should look in the developer tools, not in the “Page Source” — details in the article about weight transfer.What to use: a scenario table
| Task | What to use |
|---|---|
| Paid or affiliate link | rel="sponsored" (or nofollow) |
| Links in comments and forums | rel="ugc", optionally with nofollow |
| Cart, account area, on-site search results | meta robots noindex; keep the page crawlable in robots.txt, or the crawler never sees the tag |
| A PDF or image you do not want in search | X-Robots-Tag: noindex |
| Duplicate page with parameters | A canonical URL, not noindex — see canonical URL |
| A section that should not be crawled at all | Disallow in robots.txt (the URL can still be indexed without a description if other sites link to it) |
| Keep AI crawlers away from your content | Rules for their user agents in robots.txt — see how to block AI bots |
| A page you are promoting | None of the above: indexable, crawlable, self-referencing canonical |
What to check on your site
- The promoted page should not contain
noindex— otherwise, links lead nowhere. - It should not be closed in
robots.txt: a closed page will not be crawled by the robot, even knowing the address. - The canonical address should point to itself, not to another page.
- You can check everything at once with robots.txt check, meta tag analysis, and index checker.
How to check a link for nofollow and a page for noindex
- The page link analyzer lists every link with its anchor and a dofollow or nofollow label, plus the internal/external split. Limitation: it labels a link nofollow only if
relcontains the wordnofollow; a link with onlyugcorsponsoredshows as dofollow, so check those in the code. It reads HTML without running JavaScript, so it will not see attributes changed by scripts. - The index checker shows the robots meta tag, X-Robots-Tag header, canonical and robots.txt rules — everything that can remove a donor page from search.
- Browser developer tools (Elements tab) — for the rendered page, if the platform changes
relwith a script. - Search Console → URL Inspection — for your own pages: whether indexing is allowed and whether robots.txt blocks crawling.
A full walkthrough of checking someone else's page is in page link analysis.
Are nofollow links needed at all
Yes, as part of a natural profile: a link mass consisting entirely of dofollow looks less plausible than a mixed one. But it is not worth paying for them as if they transfer weight — that’s why we keep this flag for each platform separately.