Off-Page SEO
Chapter 04 / 08
External anchor text distribution
The mix of branded, naked-URL, generic, descriptive, and exact-match anchors on inbound links — the single most diagnostic signal of whether a backlink profile is natural or manipulated.
External anchor text — the visible label of inbound links from other sites — is the single most diagnostic signal in any backlink audit. A healthy profile shows a long-tailed mix dominated by branded references. A manipulated profile shows statistical concentration on commercial-keyword exact matches. The shape of the distribution often tells you in thirty seconds whether a campaign was earned or bought.
“Inbound anchor text was the easiest thing to manipulate before Penguin and the most heavily policed signal since. The 2026 game is making your inbound anchor profile look the way unoptimised, real-world references would look — mostly branded and naked URLs, with descriptive as a meaningful minority and exact-match as a small accent.”
The five anchor types and their healthy shares
| Type | Example | Healthy share | Risk if dominant |
|---|---|---|---|
| Branded | ‘SEOTopSecret’, ‘SEOTopSecret.com’ | 40–60% | Low — branded anchors are the most natural form |
| Naked URL | ‘https://seotopsecret.com/seo/backlinks’ | 10–20% | Low — naked URLs read as authentic citation |
| Generic | ‘click here’, ‘this guide’, ‘learn more’ | 5–15% | Low to none — but excess wastes the link signal |
| Partial-match descriptive | ‘the SEO playbook from SEOTopSecret’ | 10–25% | Low — descriptive anchors carry authentic topical signal |
| Exact-match commercial | ‘SEO agency Mexico’ | ≤2–5% | High — concentration past 10% reads as manipulation |
The shares are heuristics — Google doesn’t publish a fixed table — but they reflect what natural distributions look like across hundreds of audited profiles. The ranges are wider than you’d expect because the right number depends on your category and brand strength.
Why exact-match concentration is the killer
Exact-match commercial anchors are the only anchor type that does not arise organically at high frequency. A real journalist citing your page uses your brand name, your title, or a descriptive phrase — they don’t use the precise commercial keyword you’d optimise for. So when an algorithm sees 30% of inbound anchors hit the same commercial-keyword exact match, the only plausible explanation is that someone arranged for those anchors to exist.
Penguin in 2012 was the first major algorithm to operationalise this; the link-spam updates of 2017–2019 tightened it; the 2024 spam refinement caught the next layer of bulk-paid networks running this exact pattern. The detection has gotten progressively better, and the recovery time has gotten progressively longer.
How to audit your existing distribution
Most professional tools surface anchor distribution directly. The workflow:
- Pull the anchor breakdown. Ahrefs > Backlink profile > Anchors. Semrush has the equivalent under Backlink Analytics > Anchors. Filter to dofollow only for the cleanest signal.
- Categorize the top 50 anchors. Tag each as branded / naked URL / generic / descriptive / exact-match. The top 50 typically cover >80% of the link volume.
- Sum by category. If exact-match commercial anchors aggregate above 10%, you have a yellow flag. Above 25%, you have a red flag and probably an active or imminent algorithmic depression.
- Compare to top-ranking competitors. Pull the same breakdown for the pages currently ranking. Their distribution is your reference point — if your exact-match share is double theirs, you’re visibly out of band.
The fix — dilution, not removal
When an audit finds over-optimized anchor concentration, the instinct is often to disavow the offending links. That’s usually wrong. The right move is dilution: aggressively earn or build new links with branded, naked-URL, and descriptive anchors until the over-optimized share drops below the threshold.
The timeline is 6–12 months for most profiles. The volume target depends on the existing imbalance — you’re trying to push the exact-match share from 25% down to under 10%, which means roughly tripling the non-exact-match link count over the dilution period. The dilution links should come from the legitimate tactics in the Link Building Strategies chapter — digital PR and HARO are particularly useful here because they reliably produce branded and descriptive anchors.
When to disavow instead of dilute
Disavow is correct in two cases: (1) when the offending links are demonstrably toxic — clear PBN signatures, paid-link networks, foreign-language link farms targeting your domain — and (2) when a manual action notice has been received from Google’s Webspam team. Outside those cases, dilution is faster, lower-risk, and produces a healthier profile when complete. The toxic-link diagnosis is the subject of the next chapter — Toxic Backlinks and Disavow.
Anchor distribution by tactic
Different link-building tactics produce different anchor mixes by default. Knowing this lets you steer the distribution intentionally:
- Digital PR with original data: mostly branded + descriptive (journalists naturally use the source brand). Strong cleanup tactic.
- HARO / Qwoted source pitches: almost entirely branded + named-expert credit. Cleanup gold.
- Broken-link building: mixed — depends on the original anchor on the broken link. Often descriptive.
- Resource-page placements: typically descriptive (the resource page describes what you offer).
- Guest contributions: author-bio links are usually branded; in-body editorial links are often descriptive or naked URL.
- Bulk guest posts on networks: historically the source of exact-match concentration. Avoid; this is the pattern Penguin was built to catch.
The internal vs external split
One last clarification: the rules above apply only to external (inbound) anchors. Internal linking inside your own site follows different rules — exact-match repetition is fine and even preferred for internal anchors, because Google treats internal anchors as deliberate site-architecture signals rather than third-party endorsements. Mixing the two domains is the most common confusion in this topic. Internal linking is covered in the Internal Linking chapter.
Where to go next
The next chapter — Toxic Backlinks and Disavow — covers how to identify links that are actively hurting the profile, when disavow is the right move, and how to construct the disavow file Google will actually act on.
Common questions
Common questions
Quick answers to what we get asked before every trial signup.
There’s no single correct distribution — it varies by industry, brand strength, and the natural way your category is referenced. But the long-running rough benchmark for healthy profiles is: 40–60% branded, 10–20% naked URL, 5–15% generic (‘click here’, ‘this guide’), 10–25% partial-match descriptive, and ≤2–5% exact-match commercial. Profiles that deviate sharply from these bands — especially exact-match above 10% — start triggering algorithmic flags.
In this cluster