Follow vs nofollow links: when each passes value

A follow link passes ranking signals (the modern descendant of PageRank) and counts as an endorsement; a nofollow link carries a rel="nofollow" attribute that tells Google not to treat it as a vote. Since March 2020 nofollow is a hint, not a strict command, so Google may still pass value, but you should assume it usually does not. Both types belong in a healthy profile.

That two-sentence answer is the part most guides get wrong by burying it under 800 words of history. The nuance worth your time is when each attribute actually moves the needle, and that depends less on the tag itself than on where the link lives, why it exists, and what it does for real humans. This guide walks through the mechanics, the 2019-2020 rule change that everyone keeps forgetting, and the practical decision framework we use at ANGLE when we place and audit links on a daily basis.

What the rel attribute actually does

Every link is by default a follow link. There is no rel="follow" tag, despite what half the internet calls them. "Dofollow" is SEO slang for the absence of a blocking attribute. A link only becomes nofollow when someone adds rel="nofollow" (or its cousins sponsored and ugc) to the anchor in the HTML, like this: anchor.

Google introduced nofollow in 2005 to fight comment spam. For roughly 14 years it worked as a directive: Google would not crawl that link for discovery and would not pass any ranking credit through it. The link still existed for users, it just told search engines to look away.

The reason this matters for link building is the same reason PageRank matters: a follow link is a signal of trust that flows authority from the linking page to yours. A nofollow link removes that flow. Understanding the difference is foundational to everything in link building, which is why we treat it as a first-day concept rather than an advanced one.

The 2020 shift: hints, not directives

This is the single most misunderstood fact in the follow vs nofollow debate. On 10 September 2019 Google announced that nofollow would change from a directive to a hint, and the change took effect for crawling and indexing on 1 March 2020. Per the official Google Search Central blog, all link attributes now "work as hints" that Google can choose to honor or ignore for ranking and crawling purposes.

In plain terms: Google reserves the right to crawl a nofollow link and even pass some value through it if its systems decide the link is genuinely relevant. As Search Engine Land reported at the time, this gave Google flexibility rather than a guarantee. The practical takeaway has not changed in six years: plan as if nofollow passes nothing, but be glad when it occasionally does.

The three link attributes you need to know

In 2019 Google split nofollow into three values. They all act as hints today, but each tells Google something different about why the link exists. Using the right one protects you from looking like you are buying ranking signals.

You can combine them, for example rel="ugc sponsored" for a paid placement inside a forum. Google's official qualify-outbound-links documentation confirms that if your old paid links already use nofollow, you do not need to migrate them to sponsored, you just need at least one of the two so you are not at risk of a link-scheme action. This is the compliance side of buying links: the attribute is how you disclose money changed hands.

When each type actually passes value

Here is the decision framework. "Value" is broader than PageRank. A link can be worth chasing for ranking signal, for referral traffic, for brand entity building, or for AI-search visibility, and these do not all require a follow attribute.

Follow links: the ranking engine

A follow link from a relevant, trusted page is still the most reliable way to move rankings. It passes authority, it strengthens topical association, and combined with good anchor text ratios it tells Google what your page should rank for. But follow status is necessary, not sufficient. A follow link from a thin, irrelevant, or spammy site can be worthless or worse. Relevance and trust decide the value, not the attribute. We break down the full quality stack in what makes a good backlink, and the mechanics in how Google evaluates links.

Nofollow links: the supporting cast that still earns

Nofollow links pass value in ways that have nothing to do with the attribute:

• Referral traffic. A nofollow link in a popular newsletter or Reddit thread can send more qualified visitors than ten follow links from dead blogs.
• Brand and entity signals. Mentions on Wikipedia, major news sites, and forums, almost all nofollow, build the brand demand that correlates with stronger organic performance over time.
• AI-search visibility. A 2025 Semrush analysis cited across the industry found nofollow links correlate with AI visibility almost identically to follow links, because tools like ChatGPT and Google's AI Overviews read citations regardless of attribute.
• Profile naturalness. A site that earns links organically always accumulates nofollow links. A profile that is 100% follow looks manufactured.

That last point is why high-authority publishers default to nofollow. Per Ahrefs research, roughly 10.6% of backlinks to the top 110,000 sites are nofollow, and sites like Wikipedia, Forbes, CNN, and Entrepreneur apply nofollow to most external links. Wikipedia went fully nofollow back in 2007 to kill spam incentives. Those links still carry enormous referral and brand weight.

How to check whether a link is follow or nofollow

You do not need to trust a placement provider's word. Verify it in seconds:

1. View source. Right-click the page, choose "View page source" or inspect the element, and look at the tag for your URL. If you see rel="nofollow", sponsored, or ugc, it is not a follow link.
2. Use a browser extension. Tools that highlight nofollow links on a page let you scan a whole article at a glance.
3. Run a bulk check. For a full profile, an SEO crawler or backlink tool will label every link's attribute. Our free link tools let you inspect attributes without paying for a suite.

How we approach it when placing links

When we place editorial links on our DR55 domain through /place-a-link, they are genuine follow placements inside real editorial content, because that is what earns ranking value and survives a manual review. When a link is paid in a way that should be disclosed, the honest move is sponsored, and we are upfront that it trades ranking credit for reach and relevance.

The mistake we see most often in audits is a portfolio bought entirely as "guaranteed dofollow" from link farms. Those follow attributes are technically present and practically worthless, because Google has already discounted the sources. A handful of nofollow mentions from a trusted publication will outperform them on every metric that matters, including referral revenue. Attribute is the last thing we check, not the first. Compare the economics on our link pricing index before you assume follow always costs more for good reason.

Bottom line

Follow links are your ranking engine and should make up the bulk of the links you actively build. Nofollow links are not failures, they are the referral traffic, brand mentions, and AI citations that a natural profile accumulates and that increasingly drive visibility on their own. Tag paid links honestly with sponsored, never strip nofollow off a link just to fake a signal, and judge every link by relevance and trust first. The attribute is the smallest part of the decision.