Backlink Follow and Nofollow: What Each One Does and How to Audit the Mix You Already Have

A followed backlink is an ordinary link that Google is free to use when it works out how pages relate to each other. A nofollow backlink carries a rel value, rel="nofollow", rel="sponsored" or rel="ugc", and Google generally does not count it for ranking. Both kinds sit in every real link profile. The useful question is whether each link sits in front of the right readers, and whether its attribute describes what the link truly is.

One piece of vocabulary first. There is no dofollow attribute in HTML. A followed link is simply an anchor with no nofollow, sponsored or ugc value on it. The word survives because link sellers needed a label for what they were selling, which is covered in our piece on dofollow backlink lists.

The four states a link to you can be in

What the HTML shows	Google's stated use	Where you meet it	Counted for ranking
No rel value	An ordinary editorial link	Articles, resource pages, supplier and member listings, news coverage	Yes, Google is free to use it
rel="sponsored"	Advertisements and paid placements	Sponsored posts, paid directory entries, affiliate links, event sponsor lists	No, read as a hint
rel="ugc"	User generated content such as comments and forum posts	Blog comments, forum threads, question and answer sites	No, read as a hint
rel="nofollow"	Used when the other two values do not apply	Blanket site rules on large publishers, wikis, social profiles, press release wires	No, read as a hint

Values can be combined. Google's guidance on qualifying outbound links says you may specify multiple rel values as a space or comma separated list, so a paid link inside a comment can carry both sponsored and ugc.

The 2019 change that decides how to read a nofollow link today

In September 2019 Google introduced sponsored and ugc, and moved all three values to a hint model. The announcement is explicit about the timing: "All the link attributes, sponsored, ugc, and nofollow, now work today as hints for us to incorporate for ranking purposes. For crawling and indexing purposes, nofollow will become a hint as of March 1, 2020." The same post says of comment and user generated links that "We'll generally treat them as we did with nofollow before and not consider them for ranking purposes" (Google Search Central, Evolving nofollow).

Two practical readings follow from that. First, plan as though a nofollow link carries no ranking credit, because Google says it generally treats those links the way it always did. Second, the attribute is no longer a wall. Google's documentation notes that linked pages may still be found through other means such as sitemaps, and that keeping a page out of the index is a job for the noindex rule with crawling allowed, not for a link attribute.

How to check whether a given link is followed

Three ways, in rising order of effort.

• Inspect the anchor. Right click the link in a browser, choose Inspect, and read the rel value on the anchor tag. This is the fastest check for a single placement.
• Search the source. Open view-source on the page and search for your domain. Look at the whole anchor, not just the href, because the rel value may sit before or after it.
• Check from the command line when you are reviewing a batch of placements.

curl -s https://example.com/the-article | grep -o '<a[^>]*yourdomain\.com[^>]*>'

The Links report in Search Console lists the sites and pages linking to you, and it does not show the rel value, so the attribute check is always a manual or scripted step. Recheck your most important placements once a quarter, because a publisher can change a site wide link policy long after your article went live.

Why a nofollow link still earns its place

Chasing followed links only is the mistake I see most often, because it ignores four things a link does besides passing credit.

• It sends people. A nofollow link in a buying guide that your market reads sends better traffic than a followed link on a page nobody visits.
• It puts you in the sources buyers check. Review sites, trade directories, association member lists and community threads are often nofollow by policy, and they are also the pages a cautious buyer opens before contacting you.
• It puts you in the sources software reads. A model summarising a category reads the sentence around the link whatever the rel value says.
• It makes the profile look like a business. A profile that is almost entirely followed links from content sites is a profile somebody built. Real companies collect social profiles, forum mentions, press wire pickups and supplier listings along the way.

An audit of your own follow and nofollow mix

This takes an afternoon and gives you a list of actions rather than a number to admire.

1. Export your referring domains from whichever link tool you already pay for, plus the Search Console Links report so you have Google's own view.
2. Keep one row per domain, not per link, and record the best page that links to you from each one.
3. Classify the attribute on that best link using the checks above. Script it if the list runs past a hundred domains.
4. Add two columns of judgement: is this audience the one that buys from you, and did the link arrive because somebody chose to cite you or because somebody was paid.
5. Sort by audience, not by attribute. You now have four groups: followed and relevant, which you protect; nofollow and relevant, which you keep feeding because it brings readers; followed and irrelevant, which deserves a second look for link scheme risk; nofollow and irrelevant, which you ignore.
6. Write down the ten placements in the relevant groups whose publishers also run editorial sections, and pitch those sections. A relationship that started with a directory entry is the warmest route to a cited mention.

If you are starting from scratch rather than auditing, our explainer on what a backlink is and what makes one valuable covers the ground beneath this.

When a followed link is the thing to fix

A followed link on a paid placement is a problem for the site that sold it and for the site that bought it. Google's spam policies treat buying or selling links that pass ranking credit as a link scheme, and the documented remedy is to mark the link sponsored or nofollow. So if you sponsor a newsletter, pay for a directory listing or run an affiliate programme, ask for sponsored on those links and say so in writing when you brief a publisher. Be wary of any provider whose pitch is a count of followed links at a price per link, since the thing being sold is the thing the policy names.

Questions about follow and nofollow backlinks

Do nofollow links help rankings at all?

Plan as though they do not. Google says it generally does not consider them for ranking purposes. They help in the other ways listed above, which is why they stay on your target list.

Should I ask a publisher to remove the nofollow?

Not on a site with a blanket policy, because the answer is already written and asking spends goodwill you will want later. On a one off editorial link where the rest of the page links out normally, a polite note pointing out the inconsistency is reasonable.

What ratio of followed to nofollow links should I aim for?

The ratio is an output of how you earn links rather than a number you set. A business that gets cited, listed, reviewed and discussed ends up with a mix of both.

Does a nofollow link still get my page crawled?

It may. Google's documentation says these links will generally not be followed, and also that the linked pages may be discovered by other means such as sitemaps. Use a noindex rule when you need a page kept out of the index.

Which value should a paid placement carry?

Sponsored is preferred by Google, and nofollow remains acceptable for the same job. Use ugc for links inside comments and forum posts.

How do I keep the audit current?

Recheck the top twenty placements quarterly and the rest annually. Log the date and the attribute in the same sheet so a site wide policy change shows up as a row that changed rather than a surprise.

The link that passes nothing and still decides whether you are named

Run the same audit from a different angle and it reads differently. Ask ChatGPT, Gemini, Perplexity or Claude the buying question in your category, and the answer comes back built from pages those systems can read: community threads, reference sites, comparison articles and trade directories. A large share of those pages mark every outbound link nofollow by policy. The attribute tells Google how to treat the link. It says nothing at all about whether a model reads the sentence around it, and that sentence is what carries your name into the answer. So the forum thread you filed under "passes no credit" can be the exact source that gets you recommended, while a followed link on a page no model ever reads does nothing for that answer. AI Knows Us exists to read that second list: which pages an assistant leaned on when it answered, and whether any of yours were among them. Put the domain list you just classified beside that reading at aiknowsus.com, and the nofollow column stops looking like a write off.