Dofollow vs Nofollow: We Audited Our Own 274 Outbound Links and Found Eight Leaks

We ran the audit on ourselves before writing this. Across 256 published pages, voctos.com carries 274 outbound links to 103 domains. Thirty-nine of them point at agencies we compete with, and eight of those were passing link equity we never meant to give away.

That is the part most explainers skip. The mechanics of the two attributes take a paragraph. Knowing which of your own links are leaking takes an audit, and almost nobody runs one.

Key takeaways

• A dofollow link is just a link with no rel attribute stopping it. Nofollow is the attribute that tells Google not to pass ranking signals through.
• Since September 2019, Google treats nofollow, sponsored and ugc as hints rather than strict directives.
• Nofollow does not stop a reader clicking, and it does not remove the citation. It only stops the ranking signal.
• On our own site, 31 of 39 competitor links already carried nofollow. The eight that did not sat on five pages, and one used a full descriptive sentence as its anchor text.
• The fix was adding the attribute, not deleting the citation. Removing a source you relied on turns a sourced claim into an unsourced one.

What is the actual difference?

A dofollow link passes ranking signals to the page it points at. A nofollow link carries rel="nofollow" in its HTML, which asks Google not to pass those signals.

There is no such thing as a “dofollow attribute”. The word exists only because people needed an opposite for nofollow. A normal link is dofollow by default:

<a href="https://example.com">a normal link, passes signals</a>
<a href="https://example.com" rel="nofollow">the same link, signals withheld</a>

Google introduced nofollow in 2005 to deal with comment spam, and left it alone for 14 years.

In September 2019 it changed how the attribute works and added two more: sponsored for paid and affiliate links, and ugc for user-generated content such as comments and forum posts. All three became hints rather than directives, for ranking purposes from that announcement and for crawling and indexing from 1 March 2020. A hint means Google reserves the right to consider the link anyway.

The practical effect: nofollow is a strong signal about your intent, not a lock on the door.

When does nofollow actually help you?

In four situations, and competitor links are the one most sites miss.

Paid placements and affiliate links need sponsored, and that is a policy requirement rather than an optimisation choice. Comment sections and user profiles need ugc. Untrusted or unvetted destinations need nofollow.

The fourth is the one worth your attention. When you cite a direct competitor as a source, a plain link hands them a ranking signal and a descriptive anchor, on a page you paid to produce. You want to keep the citation and withhold the signal, and the attribute is what lets you do both.

We audited our own outbound links. Here is what we found.

On 19 September 2026 we pulled every published and scheduled post and page on voctos.com through the WordPress REST API, extracted every outbound link with its rel attribute, and grouped them by destination domain.

Thirty-one of the thirty-nine already carried nofollow, so the policy existed and had simply missed eight links.

Those eight clustered on five pages: three on one article about becoming a GEO expert, two on a Google AI Overviews guide, and one each on three others. Five pages out of 256, which is 2% of the site carrying 100% of the problem.

Two of the eight are worth describing, because they show how this happens.

The first pointed at a regional competitor’s blog post. The anchor text was a full descriptive sentence, “not a single brand produced consistent results”, which is the most valuable kind of anchor a page can hand over.

Our claim depended on that source, so deleting the link on its own would have left an unsourced assertion in the text.

The second used the destination’s bare domain name as its anchor. That is the least useful form for a reader and a clean brand signal for the recipient.

Neither link was a mistake of policy. Both were a mistake of review, which is why an audit finds things a style guide will not.

Why does this keep happening on agency sites?

Because the people who write the citations and the people who set the link policy are usually not the same, and the gap is invisible from the front end.

A writer sourcing a claim reaches for the strongest available evidence. Sometimes the strongest available evidence sits on a competitor’s site, because they published the research. That is a good instinct and it should not be trained out. What it needs is a review step that checks the destination domain before publication rather than a year later.

The audit itself takes minutes once it is written. The reason it rarely runs is that nothing in a CMS surfaces the question.

How do you run this on your own site?

Three steps, and the whole thing works on any WordPress site with REST access.

Pull every published post and page with its raw content, not the rendered HTML. That way you see the attributes as stored rather than as filtered by a plugin.

Extract every tag carrying an absolute external URL, keep its rel attribute and its anchor text, and drop anything pointing at your own domain.

Comparing agencies on how they handle this? Our answers to Seer Interactive’s 27 GEO agency questions cover the reporting side in the same detail.

My read on the nofollow debate

Most of the discussion gets stuck on whether nofollow links still carry residual value. Google says they are hints, nobody outside Google can measure the rest, and another thousand words on it will not settle anything.

What you can settle is intent. Every outbound link leaving your site is a choice about who you help, and on most sites nobody has ever made that choice deliberately.

The opposite error is just as common. Blanket-nofollowing every external link makes a site look like it has nothing behind it, and sourcing well is a trust signal to readers and to the engines parsing your page. We link out to Google’s documentation, to primary research and to news sources with no attribute at all, and that is deliberate.

The rule we run is narrower than “nofollow external links”. It is: nofollow the domains that compete with us for the same client, and link everything else normally.

Questions we get about this

“If I nofollow a competitor, am I hiding the source?”

No. The link still resolves, the reader still clicks, and the citation still appears in your text. The only thing that changes is the ranking signal. If you are worried about the appearance, name the source in the sentence as well as in the link, which is better practice anyway.

“Does nofollow hurt my own page?”

There is no evidence that carrying nofollow links damages the linking page. The older worry about “PageRank sculpting” was about hoarding internal signals, which Google addressed years ago and which is a different question from external attribution.

“Should I just remove competitor links instead?”

Only if the claim survives without them. Ours did not in one case, so we replaced the competitor source with our own published data and linked internally instead. That is the better outcome when you have the data. When you do not, keep the citation and add the attribute.

“What about links inside AI answers?”

Different mechanic entirely. An AI engine citing your page is not passing PageRank, and rel attributes on your outbound links do not control whether an engine cites you. Treat AI citation as a separate measurement rather than folding it into link reporting.

Where this actually matters

The attribute itself is a five-minute fix, and eight links were never going to move our rankings. The audit is the part worth doing.

A site that nobody has read end to end accumulates decisions that nobody actually made, and ours had been accumulating them for years before we ran the query for the first time. Most sites we look at are in the same position, including ones with a documented link policy.

If you want the same audit run across your site alongside the rest of your technical setup, the technical SEO page covers how we scope it.

Related guides

• External Links: How to Build Your Site’s Backlink Profile
• Broken Links: How to Find and Fix Them Yourself
• Sovereign AI in the Gulf: What Arabic Models Mean for How Your Business Gets Found
• JavaScript Links Are Hiding MENA Sites From ChatGPT and Claude