Backlink Profile Analyzer
Paste a backlink export from Ahrefs, Semrush or Search Console and see who links to you, with what words, whether the links are followed, and whether the profile looks earned or built.
| Domain | Links | Share | Followed |
|---|
| Anchor | Type | Count | Share |
|---|
| Page | Links | Domains |
|---|
Nothing you type here is uploaded, stored or sent anywhere. It all runs in your browser.
What a backlink profile analyzer should actually tell you
Whatever the dashboard looks like, a backlink profile analyzer answers four questions: how many different sites link to you, what words they use, whether the links are followed, and whether a handful of sites are doing all the work. Ahrefs and Semrush answer all four, but across five reports and behind a subscription. This one takes the export you already have, reads it in your browser, and puts the four answers on one screen, along with the patterns I look for first when a new client sends me their profile.
Referring domains matter more than links
The headline number in most tools is total backlinks, and it is the least useful number on the page. A blog that puts you in its footer gives you a link on every page it has: two hundred links, one editorial decision. Google has never treated two hundred copies of one link as two hundred votes, and in my experience a site-wide link behaves much like a single one. So the count I care about is referring domains, and the ratio I glance at is links per domain. Three thousand links from forty domains is a small profile with a few site-wide links in it, not a big one.
That is also why this tool counts anchors per referring domain by default. Counted per link, one footer link with a keyword anchor can make exact-match look like a third of the profile. Counted per domain, it is one site making one choice. Switch the setting and look at both; the gap between the two numbers is a finding in itself.
The anchor mix is where built profiles give themselves away
People link to you with your name, your URL, or words like "here" and "this guide". They very rarely use the exact phrase you want to rank for, because nobody talks like that. So an earned profile is mostly branded and URL anchors, with exact-match in single digits. Past 10% exact-match a profile starts to look built; past 20% it looks manipulated. Those are my working lines, not numbers Google publishes, and they are the same ones the Anchor Text Distribution Checker uses, so the two tools never disagree. Partial-match anchors, with the keyword inside a longer phrase, are a much weaker signal and I do not flag them on their own.
Followed and nofollow
Since 2019 Google treats nofollow, UGC and sponsored as hints rather than rules, but in practice those links carry little. A profile that is mostly nofollow is not in trouble; it is weaker than its link count suggests. The more telling pattern is the opposite one: hundreds of links and not one nofollow among them. Forums, Q&A sites, community threads and social profiles all hand out nofollow links, and a site that gets mentioned naturally picks some up. A perfectly followed profile is usually one somebody paid for, link by link.
How to get your export into the backlink analyzer
- Ahrefs. Site Explorer, Backlinks, Export. Either encoding works. The tool reads Referring page URL, Domain rating, Anchor, Target URL and the Nofollow, UGC and Sponsored columns.
- Semrush. Backlink Analytics, the Backlinks tab, Export. It reads Source url, Page ascore, Anchor, Target url, Nofollow and Sitewide.
- Google Search Console. Links, then Export on Top linking sites or Latest links. Search Console gives no anchor per link and no follow status, so you get domain counts only, but it is Google's own list, which neither of the others can claim.
- Anything else. A plain list of linking URLs, one per line, still gives you referring domains and links per domain.
Subdomains are folded into their parent, so blog.example.com and example.com count as one site, except on hosting platforms such as Blogspot, WordPress.com or GitHub Pages, where each subdomain really is a different person.
What the flags are, and what they are not
Every threshold behind the flags is a judgement: exact-match above 10% and 20%; one domain supplying more than 30% of all links; five or more links from one site with at least 80% of them on the same anchor and page, read as site-wide; nofollow above 50%, or at zero across twenty links or more; branded and URL anchors together under 30%; and more than half of referring domains under an authority of 10. I settled on those from looking at a lot of profiles. They are tripwires for a closer look, not verdicts.
A flag is not a reason to reach for the Disavow File Generator. For most sites the fix for a lopsided profile is more good links, not fewer bad ones, and if you want to know what one prospective link is worth before you chase it, the Backlink Value Calculator does that. The authority column is whatever your export carries, so compare a profile against itself over time in one tool rather than across tools.
If the analysis tells you what you suspected, too few sites that matter and too much of one anchor, that is the problem I am hired to fix. Here is how I price it.
Questions people ask
What is a backlink profile analysis?
It is a look at every link pointing to a site as a whole, rather than one link at a time: how many distinct sites link, which pages they point at, what anchor text they use, how many are followed, and how authoritative the linking sites are. The point is to spot patterns, such as one site carrying the profile or keyword anchors that no real person would write, that a single link never shows.
What is a good ratio of dofollow to nofollow backlinks?
There is no target ratio, and anyone who quotes one is guessing. Natural profiles almost always contain some nofollow, because forums, Q&A sites, comment sections and social profiles use it by default. I do not aim for a number; the pattern that worries me is zero nofollow across a large profile, which is what a bought profile tends to look like.
What percentage of anchor text should be exact match?
In a profile that was earned, single digits. My working lines are 10%, where a profile starts to look built, and 20%, where it looks manipulated. Those are judgements from experience, not Google figures. The Anchor Text Distribution Checker uses the same lines if you want to test a list of anchors on its own.
Why do Ahrefs, Semrush and Search Console show different backlink numbers?
Each one crawls the web independently, keeps links for a different length of time, and counts them differently. Search Console shows a sample of what Google has found, not everything it knows about. None of them is the complete list, so compare a profile against itself over time in one tool rather than one tool against another.
Are site-wide backlinks bad?
Not by nature. A partner linking you from its footer, or a friend from a blogroll, is ordinary. They count for roughly one link however many pages they appear on, so they inflate your backlink total without adding much. They become a problem when the anchor is your money keyword and the placement was paid for, which is exactly the pattern this tool flags.
How large an export can I paste?
There is no row limit in the tool. A full export of tens of thousands of rows is parsed in your browser in well under a second on a normal laptop; the slow part with very large files is the text box itself, so use the file picker for those.
Is my backlink export uploaded anywhere?
No. The page has no backend. The file is read by your browser, parsed and counted in memory, and nothing is uploaded, logged or stored. Close the tab and it is gone.
Related tools
The analysis tells you what is missing. Links are what fill it.
Most profiles I look at have the same problem: too few sites that matter, rather than too many that do not. Finding and earning those links, one editorial placement at a time, is the work I do.
See link building packages Talk to Hayk
Further reading: How I evaluate a backlink before I build it · Why most backlinks do nothing for your rankings