Anchor text ratios without over-optimizing
Anchor text is the part of link building where people either obsess over the wrong thing or ignore it entirely. Both mistakes cost rankings. Stuff your money keyword into every anchor and you build a footprint that screams manipulation. Use "click here" for everything and you waste the single strongest relevance signal a backlink carries.
The truth sits in between, and it is less precise than the spreadsheet warriors want it to be. There is no magic percentage that Google rewards. What exists is a range of natural-looking distributions, a set of patterns that reliably get links devalued, and a planning method that keeps you inside the safe zone without paralysis. This article gives you all three, plus the numbers I actually aim for when I plan anchors for a page.
- There is no single correct anchor ratio. Aim for a natural distribution: branded and naked-URL anchors should dominate, exact-match should be the smallest slice.
- The pattern that gets links devalued is not one bad anchor, it is repetition: the same exact-match phrase across many domains in a short window.
- Plan anchors per target page, not per site. A page with 8 links and 6 exact-match anchors is exposed even if your sitewide ratio looks fine.
- Internal links follow different rules: you can be far more descriptive with internal anchors because you control them and there is no manipulation incentive.
- Treat ratios as guardrails, not targets. Earned links arrive with messy, varied anchors. Replicate that mess on purpose.
On this page
What anchor text ratios actually measure
An anchor text ratio is just the breakdown of how your inbound links describe the page they point to. Tools like Ahrefs and Semrush bucket anchors into categories, and that bucketing is where most of the confusion starts. Before you can plan a ratio, you need to agree on what each category means in practice.
| Anchor type | Example (target: a page on email marketing) | What it signals |
|---|---|---|
| Branded | ANGLE, angletutoring.com | A real brand citation. The safest, most common type on natural profiles. |
| Naked URL | https://angletutoring.com/guide | Someone pasted the link. Extremely common and zero risk. |
| Generic | click here, this article, read more | Low relevance signal, but very natural. Real people link like this constantly. |
| Exact match | email marketing software | High relevance, high risk. The phrase you want to rank for, verbatim. |
| Partial match | best email marketing tools for 2026 | Relevance with variation. The workhorse of a safe profile. |
| Topical / LSI | sending campaigns at scale | Related phrasing that supports relevance without exact repetition. |
| Image / empty | (alt text or no anchor) | Counts as a link, often branded or empty. Common and harmless. |
The reason categories matter: Google does not penalize a single exact-match anchor. It looks at the shape of the whole profile and asks whether it resembles links people earn or links people buy. A profile that is 70% exact-match did not happen by accident, and the algorithm knows it.
Categorization is fuzzy
The mix that looks natural
Here is the part everyone wants: numbers. I will give them, but read the warning that follows them, because using these as rigid targets is itself a footprint.
When I audit a healthy, organically-grown site that has never bought a link, the distribution usually lands somewhere in these ranges. This is descriptive, not prescriptive: it is what natural looks like, which is what you are trying to imitate.
| Anchor type | Natural range (approx.) | Notes |
|---|---|---|
| Branded + naked URL | 45% to 65% | The majority. If this is under a third, your profile already looks engineered. |
| Generic | 5% to 15% | Real people link with 'this guide' all the time. Do not be afraid of it. |
| Partial / topical | 15% to 30% | Where most of your relevance signal should live. |
| Exact match | 1% to 8% | The smallest slice. Treat anything above ~10% sitewide as a yellow flag. |
Two things jump out of that table. First, branded and URL anchors should be the bulk of your profile. New site builders find this counterintuitive because branded anchors feel like wasted opportunity. They are not. They are the ballast that lets you place a few exact-match anchors without the profile tipping over. Second, exact-match is tiny. A site ranking page one for a competitive term often has exact-match anchor density in the low single digits, because most of its links were earned and earned links rarely use your target keyword verbatim.
The ratio is not the target
The patterns that trigger devaluation
Modern Google rarely hits you with a manual penalty for anchors. What happens instead is quieter and more common: the algorithm devalues links it does not trust, so they stop passing weight. You keep paying for links that do nothing. Here are the patterns that reliably cause that.
Exact-match repetition across domains
This is the big one. One link with the anchor "project management software" is fine. Twelve links from twelve different sites all using "project management software" within two months is a fingerprint that almost never occurs naturally. Real editors phrase things differently. When you see identical exact-match anchors clustering, that is bought links failing to disguise themselves. If you are buying placements at all, the anchor plan is where the risk concentrates, so this is exactly where to be disciplined.
Velocity spikes on commercial anchors
It is not just the ratio, it is the rate. Acquiring forty exact-match anchors in a single month to a page that had ten total links last quarter is a velocity anomaly. Spread commercial anchors out. A few per month to any given page is plenty, and it gives the rest of your profile time to grow branded and generic links that keep the proportions sane.
Anchor that does not match page intent
If a page is an informational guide and ten anchors point at it with a hard commercial phrase, that mismatch is a tell. Earned links to a guide say things like "a useful breakdown of X" not "buy X cheap." Match the anchor to what the page actually is.
DR-blind anchor planning
People assume a high-authority link can carry an aggressive anchor. Sometimes the opposite is safer: a low-quality site with an exact-match anchor looks worse than the same anchor on a real publication. Authority and anchor risk interact, and it helps to understand what a Domain Rating figure does and does not tell you before you decide how hard an anchor a given source can carry.
How to plan anchors per page (not per site)
The most common planning mistake is looking at sitewide ratios. Your site can be a healthy 4% exact-match overall while one money page is sitting at 60% exact-match because that is where you pointed everything. Google evaluates the link profile of the target page, so that page is exposed regardless of how clean the site average looks. Plan at the page level.
Here is the process I use for any page I am actively building links to.
- Snapshot the current page profile. Pull the existing anchors for that exact URL. You are budgeting against what is already there, not starting from zero.
- Decide the next 8 to 10 links as a batch. Plan in batches, not one at a time. A batch lets you balance: if you want one exact-match in the batch, the other seven or eight should be branded, URL, generic, and partial.
- Cap exact-match per batch at one, sometimes two. For a brand-new page with thin existing links, zero exact-match in the first batch is often correct. Earn branded and topical signals first.
- Write partial-match anchors that read like a human wrote them. 'a solid guide to email automation' beats 'email automation software' every time, and it still carries the keyword.
- Vary the surrounding context, not just the anchor. The sentence around the link matters. Two links with different anchors but identical surrounding sentences still look templated.
- Re-check the page ratio after the batch lands. Recompute and adjust the next batch. This is a feedback loop, not a one-time plan.
A practical example. Say a page has 12 existing links: 7 branded/URL, 3 partial, 2 generic, 0 exact-match. You want this page to rank for "freelance invoicing." Your next batch of 8 might be: 4 branded/URL, 2 partial ("invoicing tools for freelancers", "how freelancers handle invoices"), 1 generic, and 1 exact-match ("freelance invoicing"). After that batch the page has 20 links with exactly one exact-match anchor. That is roughly 5%, lumpy and believable, and the single commercial anchor sits inside a sea of brand and topical signals.
Internal anchors play by different rules
Everything above is about external backlinks. Internal links are a separate game, and the rules relax considerably. You own every internal link, so there is no third-party manipulation to detect. Google treats descriptive internal anchors as a legitimate, helpful signal about what a page is about.
That means you can and should be far more direct with internal anchors. Linking to your invoicing page with the anchor "freelance invoicing software" from a relevant blog post is not a footprint, it is good information architecture. This is one of the highest-leverage, lowest-risk SEO moves available, which is why a deliberate internal linking strategy can move rankings on its own, independent of any backlink work.
The discipline that still applies internally is relevance and variety. Do not point every internal link at one page with the identical anchor from a sitewide footer; that is more about diluting the signal and annoying users than about penalty risk. Vary anchors so each link describes the destination in the most accurate words for that context. Where the external rule is "keep exact-match scarce," the internal rule is "keep anchors accurate and descriptive."
One profile, two rulebooks
A simple anchor checklist before you place a link
When you are about to place or buy an editorial link, run these questions. If any answer is wrong, change the anchor before the link goes live, because changing it afterward is far harder.
- Is this exact-match phrase already on more than a couple of my referring domains? If yes, use partial or branded instead.
- Does the anchor match the actual intent of the target page?
- Is the surrounding sentence different from sentences around my other links?
- After this link, does the target page's exact-match share stay in single digits?
- Would this anchor look out of place if a real editor had written it unpaid?
That last question is the whole article in one line. Editors who genuinely recommend a resource almost never use your money keyword verbatim. The closer your anchors get to how an unpaid editor would phrase things, the safer and more effective they are. If you want to go deeper on the placement side of this, our guide on acquiring links without leaving a footprint covers the sourcing and disclosure decisions that sit alongside anchor planning.
Frequently asked questions
What is the ideal exact-match anchor percentage?
There is no single ideal number, and treating one as a target is itself risky. On naturally-grown sites, exact-match anchors usually sit somewhere between 1% and 8% of the total profile, with the rest being branded, naked-URL, generic, and partial-match anchors. Stay in that range, keep it lumpy rather than perfectly tuned, and never let one exact-match phrase repeat across many domains. The repetition is what gets noticed, not the percentage on its own.
Can a single exact-match anchor get my site penalized?
No. One exact-match anchor is completely normal and happens on organically-earned profiles all the time. Modern Google rarely issues manual penalties for anchors anyway. What actually happens is quieter: links it does not trust get devalued and stop passing weight. That distrust comes from patterns like the same commercial phrase repeating across many sites in a short window, not from any individual link.
Do internal links count toward my anchor ratios?
They are a separate consideration. Internal anchors do not carry the same manipulation risk because you control them, so you can be much more descriptive and even use exact-match phrasing freely as long as it is accurate and varied. The catch is that backlink tools sometimes blend internal and external anchors, which inflates your descriptive count. Always isolate external referring domains before judging whether your external profile looks natural.
How fast can I build exact-match anchors safely?
Velocity matters as much as ratio. Acquiring a burst of commercial anchors to a page in a single month, especially a page that previously had few links, is a velocity anomaly that gets links discounted. Spread them out: a small number per page per month, surrounded by branded and partial-match links that grow at the same time, keeps both the ratio and the rate looking believable.
Should higher-authority links get more aggressive anchors?
Not automatically. It is tempting to assume a strong domain can carry a hard exact-match anchor, but an aggressive anchor on a low-quality site often looks worse than the same anchor on a respected publication. Authority and anchor risk interact, so judge each source on both its quality and what the rest of the target page's profile already looks like, rather than spending your exact-match budget just because a site has a high Domain Rating.