What follow ratio measures
A link either carries an attribute asking search engines not to pass credit, or it does not. Links without one are described as followed; links marked nofollow, sponsored or ugc are the other group. Backlink tools count both and report the balance, and that balance is what people mean by follow ratio.
The attribute exists for a reason that has nothing to do with your ratio. It was introduced so that site owners could publish links they had not vetted — comments, forum posts, user profiles — without appearing to endorse them, and it was later extended so that paid and user-generated links could be labelled separately. Since then search engines have treated these attributes as hints they may weigh rather than instructions they must obey, which is a meaningful change: a nofollowed link is no longer guaranteed to be ignored.
Why follow ratio matters
Mostly as a symptom, not a measurement. A profile made up almost entirely of followed links is the odd one, because the ordinary web produces plenty of the other kind. Social posts, news sites, forums, marketplaces, directories and comment sections nofollow by default, so a business that people genuinely talk about will collect them without trying.
A profile with almost no nofollowed links usually means the links were all placed deliberately, which is worth knowing about a site you are auditing or about a supplier’s past work. That is the honest use of the number: a prompt to look at where the links came from.
Where follow ratio goes wrong
The first mistake is believing there is a correct split. There is no published guidance from any search engine on what proportion should carry the attribute, and there never has been. Any supplier quoting a target is repeating something they read on a forum.
The second is engineering it. Deliberately acquiring nofollowed links to make a bought profile look natural is effort spent on appearances rather than on anything a customer would notice, and it does not address the reason the profile looked wrong in the first place.
The third, and the most expensive, is turning down good placements because they are nofollowed. Coverage in a national newspaper, a well-read industry site or a busy marketplace listing sends real readers, builds recognition and often leads to other people linking to you in turn. Judging that by the attribute alone is judging a shop window by whether the glass is branded. The same reasoning applies to nofollow links generally.
How to act on it
Stop reporting it as a health score and start using it as a question. If the followed share looks unusually high, ask where those links came from and who placed them. If it looks unusually low and rankings are flat, look at whether anything is linking to you editorially at all, or whether the whole profile is social posts and directory entries.
When you are deciding whether a placement is worth pursuing, weigh the audience first: does anyone read this site, do those readers resemble your customers, and would the mention still be worth having if search engines did not exist? That test settles most cases. Build link building work around coverage people actually see, and let the ratio fall wherever it falls.