What Is Anchor Text Ratio?
Anchor text ratio is the distribution of anchor types across the inbound links pointing at a page or a domain: branded, exact-match keyword, partial match, and more.
Anchor text ratio is the distribution of anchor types across the inbound links pointing at a page or a domain: branded, exact-match keyword, partial match, generic, and naked URL. It describes a link profile after the fact. It is not a dial you set, and no search engine publishes a target split you can build towards.
Why the mix of your anchors matters
Anchor text is one of the few things a third party writes about your page that a search engine can read directly. When hundreds of unrelated sites describe a page the same way, that description carries weight. That is the whole reason anchors matter, and also the reason a distorted profile is a useful diagnostic.
The distortion tells you something. Real citations are messy: people use your brand name, the page title, a bare URL, or “this guide”. A profile where a quarter of the anchors are the same commercial phrase was almost certainly bought, and you should know that before an agency shows you a chart with no context. If someone is quoting you for anchor cleanup, the useful move is to see what the work should cost in the open, then ask which specific links their number covers.
How the ratio is calculated
Every tool does the same arithmetic. It pulls the anchors from its own link index, buckets them, and divides each bucket by the total. The buckets are not defined by Google; they are a convention tool vendors adopted and each one classifies borderline cases differently.
| Anchor type | What it looks like for a page about running shoes |
|---|---|
| Branded | Northgate, Northgate Running |
| Exact match | best running shoes |
| Partial match | Northgate’s guide to running shoes |
| Generic | this guide, the article, read it |
| Naked URL | northgate.com/running-shoes |
| Image | whatever sits in the alt attribute |
- Choose the unit first. Per URL or per domain. They answer different questions, and mixing them is the most common way these audits go wrong.
- Deduplicate by linking domain. One site-wide footer link repeated across 8,000 pages is one editorial decision, not 8,000 anchors. Counting raw links inflates whichever anchor sits in that footer.
- Normalise the text. Strip casing, punctuation, and surrounding whitespace before bucketing, or the same phrase lands in three categories.
- Bucket, then divide. Each category as a share of unique linking domains. That is the ratio, and it is entirely a description of the past.
- Read it against the link’s source. An exact-match anchor from a genuine review is not the same object as an identical anchor from a paid guest post, even though the chart draws them the same colour.
The final step is the one tools cannot do for you. The classification depends entirely on what counts as your brand and your keyword, which means two analysts get different numbers from the same profile. The same link export can look clean in one dashboard and alarming in another. Knowing how the words inside a link work matters more here than any percentage the dashboard prints.
The blunt version
No published Google threshold exists. Every percentage target you have been sold — 60% branded, under 5% exact match, whatever the deck said — is vendor folklore. Google has never documented a safe anchor distribution, and it never will, because publishing one would hand spammers the exact recipe to stay under it.
Where the folklore came from is easy to trace. Penguin, in 2012, hit link spam hard, and the industry reverse-engineered percentage rules from small samples of affected sites. Google’s own announcement of the 2016 Penguin update stated that it had become more granular, devaluing spam rather than demoting whole sites. Devaluation changes the diagnosis completely: the bad links stop counting, so a “toxic ratio” is mostly a description of money you already wasted.
The ratio remains useful for exactly one thing, and it is not the one it gets sold for. It is evidence about procurement. If you cannot name the person who placed an exact-match anchor, someone bought it, and the question is not what percentage it represents but whether you keep paying for more. Fix the buying process, not the pie chart, and read what genuinely earns citations before you commission a cleanup.
Example
Say a B2B software firm sees a dashboard reporting 31% exact-match anchors and panics. Deduplicated by domain, the picture changes: one integration partner links from a template across every page of its documentation, using the same product-category phrase. That is a single relationship, and by unique domain the share drops into the low single digits. Nothing needs fixing, nothing needs disavowing, and the “problem” was a counting artefact. The genuinely interesting finding sits elsewhere in the same export: nine anchors, all the same commercial phrase, from nine unrelated blogs published in one fortnight. Nobody in the company can explain who arranged them.
FAQ
What is a safe anchor text ratio?
There is no published safe ratio, and any specific figure you have been given was invented by whoever gave it to you. Profiles built from genuine citations tend to be dominated by brand names and URLs, but that is an outcome of how people cite, not a target to engineer towards.
Should I disavow links because of their anchors?
Almost never on anchors alone. Google’s guidance is that most sites never need a disavow file, and modern spam handling devalues rather than penalises. If you paid for the links yourself and the arrangement is ongoing, stopping the purchases matters far more than uploading a text file.
Why do two tools report different ratios for my site?
Because each one crawls a different link index and applies its own classification rules. One counts links, another counts domains. One reads a brand-plus-keyword anchor as branded, another as partial match. Neither is authoritative, so compare a single tool against itself over time rather than comparing two tools.
Related terms
- Anchor Text — the individual unit this ratio aggregates, and where the real signal lives.
- SERP Volatility — the ranking movement wrongly blamed on anchors most often.
- Service Area Business — a case where local citations, not anchors, drive most link value.
If a proposal quotes you a target anchor percentage, ask where the number is published. There is no answer, because there is no source, and the chart is easier to sell than the procurement conversation underneath it.