Anchor Text Diversity Auditor Free Tool

Paste your inbound links — internal links or backlinks — one per line as target-url | anchor-text to audit anchor-text concentration across your site at once. For every page that receives 2 or more links, the tool works out which single anchor is used most and what share of that page's links use it. If any one anchor accounts for more than 60% of a page's links, the page is flagged for over-optimization — a classic exact-match anchor spam signal search engines can penalise. This is a read-only report; nothing is fetched or changed.

Enter one link per line as target-url then anchor-text

About Anchor Text Diversity Auditor

What Anchor Text Is and Why It Matters

Anchor text is the visible, clickable words of a hyperlink, the part a reader actually sees and clicks, as opposed to the URL sitting underneath it. In <a href="https://example.com/running-shoes/">best running shoes for flat feet</a>, the anchor text is "best running shoes for flat feet." That's the string a person reads before clicking, and it's also a string search engines read while trying to understand what the linked page is about.

Google's own Search Central documentation describes anchor text as one of the signals used to understand a page's topic and relevance, on top of the page's own content. A link with descriptive, accurate anchor text tells both readers and crawlers something real about the destination. A link that just says "click here" or "read more" tells them almost nothing beyond "there's a link here." Neither is against any rule on its own. Both are extremely common on the open web, on real sites written by real people with no coordinated strategy.

That last point (no coordination) is exactly what makes anchor text useful as a signal in the first place, and exactly what makes a heavily skewed anchor profile worth a second look. When dozens or hundreds of independent websites link to the same page, nobody expects them to agree on the exact same three-word phrase every time. Real link profiles are messy by default. This tool exists to measure that messiness (or the lack of it) for any list of links you paste in.

Anchor text is usually grouped into a handful of recognizable types, and a healthy link profile normally contains a mix of several of them rather than leaning on just one:

A page's anchor profile doesn't need to hit some exact ratio across these five types to look healthy. What it needs is variety, evidence that the links accumulated from different sources with different habits, rather than one phrase repeated by design.

Why Anchor Text Over-Concentration Is a Risk

A naturally built link profile has varied anchor text, because the people linking to a page are varied too. Some link with the brand name. Some link with a generic phrase like "this article" or "click here." Some paste the bare URL. Some use a partial match ("running shoe guide" instead of the exact target keyword "best running shoes"). Some use the exact keyword phrase, because sometimes that genuinely is the most natural, accurate way to describe the link. All of these show up in a healthy profile, mixed together, because nobody planned the wording in advance.

Over-concentration is the opposite pattern: one exact-match phrase accounts for most or nearly all of a page's inbound anchor text. That's a much harder thing to produce by accident at scale. It usually means one of two things happened: either the links were built deliberately with a single target keyword in mind (a common, if outdated, link-building tactic), or a smaller number of link sources are reusing the same template, plugin default, or brief across many placements.

Google's 2012 Penguin update specifically targeted manipulative link-building patterns, and exact-match anchor text concentration was one of the clearest fingerprints of the schemes it was built to catch: paid link networks, low-quality guest post campaigns, and directory spam that all pushed the same keyword phrase toward the same target page, over and over. Penguin has since been folded into Google's core ranking algorithm and runs continuously rather than as a periodic update, but the underlying signal it was built around hasn't gone away. A page whose backlinks are overwhelmingly one exact phrase still reads, to an algorithm looking for that pattern, as a page that was built up rather than earned.

None of this means a single skewed page automatically triggers a penalty. Google has been consistent that manual actions and algorithmic distrust generally follow patterns, not isolated incidents, and that legitimate variation in anchor text (including plenty of exact-match anchors, in moderation) is completely normal. But a page where one exact phrase dominates the anchor profile is a page worth checking, whether you're proactively auditing your own link building or trying to explain a ranking drop you can't otherwise account for.

There's also a difference worth naming between algorithmic distrust and a manual action. Most of what Penguin-descended systems do today happens automatically and continuously, adjusting how much weight a link passes rather than flatly penalizing a whole site. A manual action, by contrast, is a human reviewer at Google looking at a site and applying a targeted action, usually reserved for clear, large-scale link schemes rather than a single page with a skewed but plausible-looking profile. An over-concentrated anchor pattern is far more likely to quietly reduce how much a set of links helps than to trigger a manual review on its own, though a large enough or clearly artificial-looking pattern can absolutely contribute to one, especially alongside other spam signals on the same links.

The 60% Threshold: A Practical Heuristic, Not an Official Google Number

This tool flags a target page as over-optimized when its single most-used anchor text accounts for more than 60% of that page's total inbound links in whatever list you paste in. That number deserves a plain, direct explanation of where it actually comes from: this tool's own practical, conservative flag point, not a percentage Google has ever published anywhere.

Google's public guidance on anchor text (its Search Central link-scheme documentation and public statements from members of its Search Relations team) consistently describes wanting natural, varied anchor text and warns against manipulative link schemes that use exact-match keyword anchors as a ranking shortcut. It has never attached an exact percentage to that guidance. There's no official line that says "under 40% is safe, over 60% is risky." Anyone who tells you Google has published a hard cutoff is repeating a number that doesn't actually exist in any Google documentation.

So why 60%? Because a threshold has to be set somewhere for a tool like this one to be useful at all, and 60% is a defensible, commonly cited rule of thumb among people who do link-building and technical SEO audits for a living: high enough that it won't flag ordinary, healthy variation (a page can easily have its single most common anchor sit at 30 or 40% and still look completely natural), low enough that it catches the pattern that actually matters, one phrase dominating almost everything.

Treat the flag as a prompt to look closer, not as a verdict. A flagged page with 65% concentration built from ten links accumulated over three years of normal, unplanned linking reads very differently from the same 65% built from ten links that all landed within the same guest-post campaign in the same month. The tool can show you the percentage. It can't see the context behind each individual link, so that part is still on you.

Why a Target Needs at Least 2 Links to Be Audited

Any target URL with only one inbound link in the list you paste in is skipped entirely, not shown in the results at all. This isn't a limitation to work around, it's a direct consequence of what "concentration" actually means.

Concentration is a statement about distribution: how a total is split across multiple parts. A single link has nothing to be distributed against. If a page has exactly one inbound link and that link's anchor text is "best running shoes," that anchor is trivially 100% of the page's link total, but the number is meaningless as a diversity signal, because there's no second data point it could have been diverse (or not diverse) relative to. One data point is not a distribution, and reporting a percentage for it would just be a confusing way of restating "there is one link here."

The practical effect: if you paste in a list where most target pages have exactly one link each, expect most of them to disappear from the results, and expect that to be correct behavior rather than a bug. The tool is designed to answer "is this page's link profile lopsided," and that question only has an answer once there's more than one link to be lopsided about.

How the Anchor Text Diversity Auditor Works

Paste one link per line into the box above, formatted as target-url | anchor-text, with the target URL first, a pipe character, then the anchor text that was used to link to it. The target must be a full, absolute URL starting with http:// or https://. The tool accepts up to 20,000 lines and counts up to 5,000 valid links; anything beyond either cap is noted, not silently dropped.

Once submitted, the tool groups every valid line by its exact target URL, then, within each group, normalizes the anchor text for comparison purposes only: leading and trailing whitespace is trimmed, runs of internal whitespace collapse to a single space, and the comparison is case-insensitive. That means "Best Running Shoes," "best running shoes," and "best running shoes" (extra spaces and all) all count as the same anchor group for a given target. When the results display an anchor group, they show the exact text and capitalization from whichever line used that normalized form first, in the order you pasted your links, not a lowercased or reformatted version.

For every target with 2 or more valid links, the tool finds the largest anchor group and calculates its share of that target's total links, rounded to one decimal place. A target is flagged as over-optimized only when that top share is strictly greater than 60%. A target sitting at exactly 60.0% is not flagged; the rule is "more than 60," not "60 or more."

If two anchor groups tie for the largest count, the tie is broken by whichever group's anchor appeared earliest in your pasted input, not alphabetically and not by anchor length. For example, if a target has 4 links total and 2 different anchors each appear exactly twice, the tool reports whichever of those two anchors was pasted first in your list as the "top anchor," even though both groups are mathematically tied at 50%. The tie-break only decides which text gets displayed, it doesn't change the reported percentage or the over-optimized flag itself, which is based purely on the number, not on which group technically "won."

The table below lists every field the tool returns for a flagged or unflagged target, so the output is easy to read at a glance without needing to guess what a column means:

Worked Example: Two Targets Run Through the Tool

The two illustrative examples below aren't pulled from a live site or a real client audit, they're a small hand-worked walkthrough so the math above is easy to follow with concrete numbers. Say this is what you pasted in:

For the first target, lines 1 through 4 all normalize to the same anchor group ("best running shoes," case and whitespace ignored), so that group has a count of 4. Line 5 is a different anchor ("running shoe guide"), count of 1. Total links for this target: 5. Top anchor share: 4 divided by 5, times 100, equals 80.0%. Since 80.0 is greater than 60, this target is flagged as over-optimized, and the reported top anchor text is "best running shoes" (the exact capitalization from line 1, the first occurrence of that normalized form).

For the second target, all four anchors are different from each other: "trail running tips," "click here," "trailtips.example.com," and "this guide." Each anchor group has a count of 1. Total links: 4. Top anchor share: 1 divided by 4, times 100, equals 25.0%. Since 25.0 is not greater than 60, this target is not flagged, and it reads as a healthy, varied anchor profile, one branded-ish phrase, one generic call-to-action, one naked URL, one generic pointer phrase, exactly the kind of unplanned mix a real, organically built link profile tends to produce.

What about a target that lands in between, say a top anchor share of 45%? The tool won't flag it, since 45 is well under 60, but it's still worth a glance if the total link count is small. Three links where two share the same anchor (66.7%, flagged) is a very different situation from thirty links where thirteen share the same anchor (43.3%, not flagged): the second case has far more independent data points behind it, so the same rough percentage carries a lot less weight as a signal. The percentage alone doesn't capture sample size, which is exactly why the "Total links" column sits right next to "Top anchor %" in the results table rather than the percentage being reported on its own.

Common Causes of Accidental Over-Optimization

Very few site owners set out to build an obviously over-optimized anchor profile on purpose. Most flagged pages get there through habits that made sense in isolation but stacked up into a skewed pattern over time. The most common patterns:

• Aggressive guest-post or PR campaigns built around one target keyword. When an outreach campaign or a paid placement service is briefed with a single target keyword phrase, that same phrase tends to get reused as the anchor text across dozens of placements, because the brief never varied and nobody along the way thought to change the wording.
• Internal linking templates and plugins with a fixed default anchor. Many CMS "related posts" widgets, breadcrumb systems, and internal-linking plugins insert the exact same anchor text (often the target page's own title or focus keyword) every single time they link to a given page. Run purely on autopilot across a large site, this alone can push a page's internal-link anchor concentration well past 60% without a single external link involved.
• Directory and citation submissions using the business name as anchor. Local business directories, citation-building services, and NAP (name-address-phone) listings almost always link using the exact business name as anchor text, over and over, across dozens of directories. This is genuinely different from the first two causes and usually isn't a red flag on its own: directory-style citations are expected to use a consistent business name, that's the whole point of a citation, and search engines generally treat this category of link differently from editorial content links, where varied wording is the natural expectation. If a flagged target's high concentration traces mostly to directory and citation links rather than editorial guest posts, that's a much lower-concern flag than the same percentage built from content links.
• A single big backlink source with unusually high per-page link counts. Occasionally one site (a partner, a syndication network, a widget embedded on many pages) links to the same target many times using the same anchor, simply because that's how its own templated linking works. A handful of large sources like this can dominate a target's anchor mix even when every other backlink to that page is naturally varied.

How to Use This Tool

This is the literal, step-by-step process for running an audit, separate from the grouping and scoring explanation above.

1. Gather your link list. Good sources: a backlink export from a paid tool like Ahrefs or SEMrush, a CMS export of internal links, a spreadsheet you've been keeping during a link-building campaign, or a manual list of links you've collected while investigating a specific page.
2. Format each link as one line: the target URL, a pipe character, then the anchor text that was used. Include the protocol on the target URL (https:// or http://), a bare domain or relative path won't be accepted as valid.
3. Paste the full list into the text box at the top of this page. Multiple links pointing at the same target are expected and required, that's exactly what the grouping logic needs to work with.
4. Complete the verification step if one is shown, then submit the form.
5. Read the summary line first to confirm how many valid links were counted and whether anything was skipped or truncated.
6. Scan the results table for any target marked over-optimized. Flagged targets are sorted to the top so you see them first.
7. For each flagged target, check the top anchor text and its percentage, then think about where those links actually came from (see the causes section above) before deciding whether it's a real concern or an expected pattern like directory citations.

Fixing an Over-Optimized Anchor Profile

There's no undo button for anchor text on links that already exist and that you don't control, most inbound links to a page were placed by someone else's site, not yours. What you can control is what happens going forward, and, in the case of internal links, what your own site is doing today. Practical fixes, roughly in order of how directly you can act on them:

• Diversify anchor text choices in every future outreach or guest-post brief. Stop specifying one exact keyword phrase as the required anchor. Give writers and outreach partners a short list of acceptable variations (branded, partial-match, generic, and the exact phrase as just one option among several) so future links land more naturally.
• Vary internal link anchors by hand instead of relying on an automated default. Where a CMS widget or plugin always inserts the same anchor for a given page, consider overriding it in your main content links with a naturally varied phrase instead of leaving every internal link on the exact same wording.
• Treat directory and citation anchors differently from editorial anchors when judging a flag. If a flagged page's concentration comes mostly from consistent business-name citations, that's not something you need to (or generally should) go fix, it's the expected behavior of that link category. Focus fixing effort on editorial and guest-post links instead, where wording variation is actually the natural pattern.
• Slow down or diversify a paid link-building campaign that's currently in progress. If you're actively running an outreach or guest-post campaign and this tool flags the exact target it's building links to, that's the moment to adjust the brief for remaining placements rather than after the campaign wraps and the pattern is already locked in.
• Re-run this tool periodically on your highest-value pages. Anchor concentration builds up gradually. A quarterly or post-campaign check on your top landing pages catches a drifting pattern while it's still a handful of links, well before it becomes dozens.
• Keep a running spreadsheet of anchor text used per placement while a campaign is active. It's far easier to vary wording deliberately while you're still choosing it, one placement at a time, than to look back at 40 completed placements and realize they all used the same three words.
• Prioritize link quality over anchor-text engineering in general. A page earning links because people genuinely find it useful will naturally end up with varied anchor text, since nobody's coordinating the wording. Chasing a "perfect" anchor ratio by adding more artificial links to dilute an existing concentration just adds a second manipulative pattern on top of the first one.

What This Tool Doesn't Do (and What to Pair It With)

Being upfront about the limits of a report-only auditor matters as much as explaining what it does check. This tool does not crawl the web or discover backlinks on its own, it only analyzes the exact target-url | anchor-text lines you paste in. If a real backlink exists somewhere but you didn't include it in your list, it simply isn't part of the calculation.

It also doesn't check whether a link is nofollow, sponsored, or ugc, doesn't evaluate the quality or authority of the linking domain, and doesn't verify that a target URL actually returns a working page (a link pointing at a page that now 404s still counts toward that target's link total and anchor concentration, since the tool has no way of knowing that from the pasted text alone). None of that is a flaw in the anchor-concentration math specifically, it's simply outside what a paste-a-list tool focused on one question can reasonably check.

In practice, this means the auditor works best as one step in a small toolchain rather than a single all-in-one report: pull your link list from a backlink index or CMS export first, run it through here to flag concentration issues, then cross-check anything flagged against the actual linking pages if you want the full picture, including link attribute type and whether the source page still exists at all.

Related Tools

A few other free tools on this site pair naturally with an anchor-text audit, depending on what you're working on:

• If you're building the link list you're about to paste into this tool from scratch, the Backlinks Generator helps put together a starting set of link placements to track and vary from day one.
• Once you know which anchors and targets you're working with, the Backlink Indexer is a natural next step for making sure those links are actually indexed and counted, not just placed.
• If a flagged target's links are old enough that some of them might no longer even resolve, running the Broken Links Checker confirms which of those links are still live before you spend time worrying about their anchor text at all.

Frequently Asked Questions

What counts as "over-optimized" anchor text?

A target page where one anchor text (after normalizing case and whitespace) accounts for more than 60% of that page's total inbound links, as pasted into this tool. See "The 60% Threshold" section above for exactly why 60% and why it's a heuristic, not an official number.

Is the 60% cutoff an official Google guideline?

No. Google has never published an exact percentage for anchor text concentration. This tool's 60% flag point is a practical, conservative rule of thumb, explained in full in the threshold section above, not a rule sourced from Google itself.

Why don't single-link target pages show up in the results?

Concentration is a distribution concept, it describes how a total splits across multiple links. A single link has nothing to be distributed against, so there's no meaningful percentage to report. See "Why a Target Needs at Least 2 Links" above.

Can I paste in internal links, backlinks, or both at once?

Yes. The tool doesn't care about the source of a link, only the target URL and anchor text pair. Internal links, external backlinks, or a mixed list all work the same way, though mixing the two in one pass means a flagged target's concentration could be coming from either source, so it's worth checking the "Common Causes" section above to figure out which.

My directory citations are flagged. Should I fix them?

Usually not. Directory and citation links using a consistent business name are a different, generally expected pattern, unlike editorial content links where varied anchor wording is the natural signal. See the directory-citation point in "Common Causes of Accidental Over-Optimization" above before treating a citation-driven flag the same as a guest-post-driven one.

Does this tool crawl my site or discover backlinks on its own?

No. It only audits the exact links you paste in, it doesn't crawl the web or discover backlinks itself. Pull your link list from a backlink tool, a CMS export, or your own tracking sheet first, then paste it in. See "How to Use This Tool" above for the full step-by-step process.