A backlink check tells you which sites link to a domain, from which pages and with what text: for your own site you can do it at no cost with Google Search Console, for any other site with a backlink checker such as Ahrefs, Semrush, Majestic or SE Ranking. The number these tools return needs cleaning before you use it, though: in our test on 11 Italian sites on 10 October 2026, 56.8% of referring domains had a spam score of 55 or more, and only 3.1% were Italian (.it) domains.
It always goes the same way.
You put your domain into a free checker and a nice, reassuring round number comes up. 898 domains linking to you.
Great, isn't it?
Depends who they are. On that site, one of the LED wall manufacturers on page one of Google Italy for "ledwall", 732 of the 898 domains were spam and not one was Italian.
The mistakes we see most often in audits are three:
- comparing the total number of backlinks between two sites, as if 900 beat 300 by definition;
- trusting a single tool, when every vendor has a different index and counts in a different way;
- rushing to disavow because a tool has coloured half the profile red.
We measure backlinks every month on the projects we work on. For Macropix, an Italian LED wall manufacturer, the category keyword went from position 88 to 2 and share of voice on the cluster reached 25.55%, ahead of Amazon at 18.08% (Semrush data for Google Italy; the case is told in the guide to SEO for manufacturing).
And what has that got to do with checking backlinks?
It matters because, while Macropix was climbing, the backlink counts of many competitors were growing faster than its own. The highest number wasn't the strongest profile. And on visilay.com, over the past few weeks, we've been watching the same thing from the inside: you'll find it further down, with the figures.
Here you'll find:
- which tools to use at no cost and what each one hides from you;
- the test on 11 Italian sites, with the full table;
- how to read a backlink report in five steps;
- when "toxic" links really matter, according to what Google has put in writing;
- how to check competitors' backlinks without copying the wrong ones.
Let's go!
How to check backlinks for free (and what each tool hides from you)
For your own site the most reliable source is Google Search Console, because it shows the links Google has actually found. For other people's sites you need the indexes of SEO tools, which crawl the web on their own account and return numbers that differ from one another.
Every free version has a specific limit. We collected them from the official pages, consulted on 10 October 2026:
| Tool | Which sites | Limit of the free version | Index size stated by the vendor |
|---|---|---|---|
| Google Search Console | only your own, verified | a sample; tables of 1,000 rows, export up to 100,000; doesn't say which links are nofollow | the links Google has found |
| Bing Webmaster Tools | yours and any other site | backlinks of similar sites included | not stated |
| Ahrefs Webmaster Tools | only your own, verified (unlimited) | 1,000 backlinks visible at a time | over 15 trillion live backlinks, updated every 15 minutes |
| Semrush (free account) | any site | 10 queries a day | over 43 trillion backlinks, about 390 million referring domains |
| SE Ranking | any site | 5 checks a day | 2.9 trillion backlinks |
| Majestic | any site | full free analysis only on a different site every few days | 186 billion URLs crawled in the last 120 days (Fresh Index) |
| Ubersuggest (Neil Patel) | any site | "a limited sample", no number stated | not stated |
Two things the table tells you.
First: for your own site you don't need to pay anything. Search Console plus Ahrefs Webmaster Tools cover routine checks; Bing Webmaster Tools adds, at no cost, competitors' backlinks without you having to verify them. Very few people use it, which is a shame.
Second: index figures can't be compared. Semrush claims nearly three times as many backlinks as Ahrefs, and Majestic counts URLs, not links. Each applies its own filters. You can see it clearly in the studies vendors run on themselves: in Semrush's 2021 study of 10,000 domains, Semrush wins; in DataForSEO's, on 10 sites, DataForSEO comes second and Ahrefs third. We haven't found an independent study.
We've compared each tool's UK database, where the vendor publishes it, in the guide to Ahrefs alternatives. If the free side is all you need, the rest is in our overview of free SEO tools.
How much of the count is noise (test on 11 Italian sites, 10 October 2026)
More than half, in our sample. 56.8% of the referring domains of the sites we analysed had a spam score of 55 or more out of 100, and .it domains made up 3.1%.
The test was run on the Italian market, where we work, and here's the method so anyone can repeat it in their own country. We searched Google Italy on mobile for three queries from sectors we know from our projects: "ledwall" (LED wall), "ricambi vespa" (Vespa spare parts) and "impianti fotovoltaici" (solar PV systems). From page one we took the sites of companies and sector operators, leaving out marketplaces, big brands, public bodies and encyclopaedias: 12 sites, one of which had no data in the index, so 11. For each one we queried the DataForSEO backlink index, counting referring domains, those with a spam score of 55 or more, and those with a .it extension. In a UK version of the test, the equivalent of the .it column would be .uk domains. We're not publishing the sites' names: they aren't our clients and the point isn't to single anyone out.
| Site (sector) | Referring domains | With spam score ≥ 55 | % spam | .it domains | % .it |
|---|---|---|---|---|---|
| LED wall 1, e-commerce | 145 | 59 | 40.7% | 1 | 0.7% |
| LED wall 2, sector blog | 56 | 17 | 30.4% | 18 | 32.1% |
| LED wall 3, audio-video e-commerce | 93 | 43 | 46.2% | 11 | 11.8% |
| LED wall 4, manufacturer (.com) | 898 | 732 | 81.5% | 0 | 0% |
| Vespa parts 1 | 363 | 194 | 53.4% | 14 | 3.9% |
| Vespa parts 2 | 261 | 166 | 63.6% | 2 | 0.8% |
| Vespa parts 3 | 405 | 199 | 49.1% | 14 | 3.5% |
| Vespa parts 4 (.com) | 401 | 199 | 49.6% | 5 | 1.2% |
| Vespa parts 5 | 373 | 191 | 51.2% | 7 | 1.9% |
| Solar PV 1, e-commerce | 256 | 62 | 24.2% | 11 | 4.3% |
| Solar PV 2, information portal | 114 | 49 | 43.0% | 20 | 17.5% |
| Total, 11 sites | 3,365 | 1,911 | 56.8% | 103 | 3.1% |
| visilay.com, for comparison | 865 | 670 | 77.5% | 10 | 1.2% |
Four things the test tells you.
The median is 49.1%. Take out the most inflated site, the manufacturer with 732 spam domains out of 898, and the total drops to 47.8%. It isn't the effect of one extreme case: in a normal profile, almost one referring domain in two is noise.
The profile with the most referring domains is the one with the fewest links from its own country. Zero .it out of 898. Anyone who looks only at the total ranks it first; anyone who looks at the make-up ranks it last.
Spam arrives in waves, and hits everyone at once. Four of the five Vespa parts sites got a link from the same .xyz domain on the night of 10 October, between 3:00 and 3:07 (UTC). None of them asked for it.
The sites with the most links from their own country are the ones that publish content. The LED wall blog (32.1% .it) and the solar portal (17.5%) get cited by other Italian sites; the e-commerce sites almost never. It's practical confirmation of what we say in the guide to link building: the links that count come from content worth citing.
An older data point points the same way. In 2019 we analysed the profile of an Italian retail chain: 341 domains out of 542 classed as high risk, 63%, but those domains sent only 1,091 backlinks out of 5,622, 19%. Lots of spam domains, few links each. Counting domains inflates the noise more than counting links does.
The visilay.com case: 413 more referring domains in five days (and where they come from)
On 5 October 2026 visilay.com had 452 referring domains in the DataForSEO index. On 10 October it had 865. We described the start of this wave in the article on negative SEO; here one detail matters, and it concerns backlink checkers themselves.
430 of the 865 domains have the word "checker" in their name. That's 49.7%.
Names like backlinkcheckerbest.site, highdapachecker.shop, blogcheckeronline.website. They're "SEO analysis" portals that automatically publish profiles of thousands of domains, each with a link to the domain described. None of them chose us: the count grows without a single site having linked to us for a reason.
So does the backlink check inflate the number it's measuring?
Partly, yes: a slice of that number is produced by the checkers' own pages. The other figures from 10 October: 670 domains out of 865 with a spam score of 55 or more, 10 .it domains in total. The most frequent extensions among incoming links are .site, .store, .space and .website, all ahead of .com.
Did we do anything? No. No disavow, no removal requests. The reason is two sections further down.
How to read a backlink report in 5 steps (and which numbers to ignore)
You read a backlink report starting from referring domains, not from the total number of links, and you strip out the noise before drawing conclusions. This is the order we use in SEO audits.
1. Look at referring domains, not backlinks
A site that links to you from the footer of 533 pages counts as 533 backlinks and a single domain. It actually happened in our sample: the LED wall blog gets 533 links from one Italian site, all in the footer. The number that tells you something is the number of domains.
2. Remove the obvious spam
Filter by extension (.xyz, .site, .store, .shop), by words in the name (checker, backlink, seo, rank) and by the tool's spam metric. Remember that spam score, toxicity score and the like are vendor metrics: Google doesn't read them. They're there to clean up the view, not to tell you what Google thinks.
3. Check which pages receive the links
If 90% of the links point to the homepage, category pages and content have no outside support. On an e-commerce site it's the most common case, and it explains why a product page struggles even on a domain with a good domain authority score.
4. Read the anchor text
A natural profile is dominated by the brand name, bare URLs and generic phrases. Lots of anchors identical to the commercial keyword, from different sites, are the signal to look at carefully, because they point to bought links. We explain how to write and read them in the guide to anchor text.
5. Track new and lost links
A spike of new links all on the same night, like the one on the Vespa parts sites, is almost always automated spam. A link lost from a trade publication or an industry association, on the other hand, is worth an email asking for it to be restored.
And nofollow backlinks? Keep them in the count, but separately. Search Console doesn't even distinguish them, as Google says in the help page for the report; third-party tools do.
Toxic backlinks: when to really worry (hardly ever, according to Google)
A "toxic" backlink is a link an SEO tool classes as risky; for Google the category doesn't exist. John Mueller said so at Search Central Live in New York, as Search Engine Journal reported in April 2025: "We don't have a notion of toxic backlinks internally", and the disavow tool is for sites that bought links and got a manual action: "It's not a part of normal site maintenance".
The official documentation on disavowing links is just as clear: "In most cases, Google can assess which links to trust without additional guidance". The tool is for when you have a considerable number of spammy, artificial or low-quality links and a manual action, received or likely. The file accepts up to 100,000 lines and 2 MB, and processing takes a few weeks.
The problem is that the tools push in the opposite direction. Semrush rates a profile's overall toxicity as "High" when more than 10% of its backlinks are toxic. With the figures from our test, all 11 sites would land in the high band. Including the sector blog with 18 links from Italian sites out of 56.
And this is where a backlink check gets expensive. Uploading a disavow file covering half your profile risks throwing away good links too, the ones the tool coloured red because of an odd extension or a young domain.
When it really is needed: an unnatural links warning in Search Console, links bought in the past, paid guest post campaigns. You'll find how to respond step by step in the guide to the unnatural links warning, while manual actions in general are covered on the page about Google penalties.
Is a tool telling you half your backlinks are toxic?
You've just seen that, by the tools' thresholds, all 11 sites in our test would come out as high risk. Before you touch the disavow file, let's read your profile together in Search Console and in two different indexes: which links to ignore, which to defend and which pages on your site need links they aren't getting today.
Book a strategy consultationHow to check competitors' backlinks (what to copy and what not to)
Checking a competitor's backlinks is about finding the sites that link to them and not to you, and that could link to you too. With Bing Webmaster Tools you can do it at no cost; with Semrush, Ahrefs or SE Ranking you can do it across several competitors at once, with the comparison called link gap or link intersect.
The test gives a practical rule: of the 3,365 referring domains of the sites analysed, the ones worth looking at are a minority. Trade associations, industry publications, suppliers and distributors, local blogs that review products. You can get those too, with a phone call or with better content.
Ignore the rest. Copying a competitor's profile, spam included, means buying the same links Google discounts automatically.
On Sunscape, an agrivoltaic developer in Italy, we went from 1 to 38 tracked keywords on page one in six months by working on a region-by-region page structure and on content. It's a useful reminder: checking competitors' backlinks tells you what's missing outside your site; it doesn't replace what's missing inside it. We've set out where links carry weight and where they don't in the guide to off-page SEO, and for earning them cleanly there's digital PR.
Which backlink checker to choose (based on how many sites you check)
The choice depends on how many domains you need to check and how often. For a single site, your own, the no-cost combination is enough; you move to a paid plan when you have three or more competitors to follow.
| Situation | What to use | Cost |
|---|---|---|
| Checking your own site once a month | Search Console + Ahrefs Webmaster Tools | none |
| One or two competitors, now and then | Bing Webmaster Tools + free Semrush account (10 queries a day) | none |
| Vetting a site before paying for an article on it | Majestic or SE Ranking's checker (5 checks a day) | none, or Majestic's entry plan |
| Several sites and competitors, every week | Ahrefs or Semrush on subscription | from $129 a month for Ahrefs Lite (about £98) to $139 for Semrush's SEO plan (about £105), monthly billing |
The real cost, plan by plan and in pounds, is on the pages about Ahrefs pricing and Semrush pricing; the head-to-head is in Semrush vs Ahrefs. If you already use Ubersuggest, its backlink section is assessed in the guide to Ubersuggest alternatives.
One piece of advice we always give: use two sources, not one. Search Console to know what Google sees, a third-party index to see competitors. When the two numbers don't match, and they almost never do, the difference is mostly noise.
And keep things in proportion. Gary Illyes, at Pubcon Pro in Austin in September 2023, said that in his view links aren't among the top three ranking factors (as reported by Search Engine Land). The most quoted Backlinko figure, that the first result has 3.8 times the backlinks of positions 2 to 10, comes from 11.8 million Google results in a study first published in 2016: a correlation from a decade-old sample, not a rule for Google UK.
Frequently asked questions about checking backlinks
Yes. For your own site use Google Search Console and Ahrefs Webmaster Tools, both available at no cost once you've verified the site. For other sites, Bing Webmaster Tools shows the backlinks of any domain, a free Semrush account allows 10 queries a day and SE Ranking 5 checks a day.
No. Google says the Links report is a sample, not a complete list: tables stop at 1,000 rows, exports go up to 100,000 and it doesn't say which links are nofollow. It's still the source closest to what Google actually sees.
No, in almost every case. Google says it has no notion of toxic backlinks and can ignore spam links on its own. Disavowing is for when there are lots of artificial links, usually bought ones, and a manual action has been received or is likely.
It depends on the site you're checking. For your own, the most reliable is Search Console, because it shows the links Google has found. For competitors no index is complete: Ahrefs, Semrush, Majestic and SE Ranking report different numbers and count in different ways, so it's worth comparing at least two.
It depends on how much work you do on links. For a site that isn't doing link building, a monthly check in Search Console is enough. During a digital PR campaign, check new and lost links weekly, to make sure the ones you've earned stay online.
No, not on their own. In our test the site with the most referring domains, 898, had 81.5% spam domains and no links from sites in its own country. What counts is how many relevant, trustworthy domains link to the right pages, not the total.
In short:
Checking your own site's backlinks costs nothing, with Search Console and Ahrefs Webmaster Tools.
In our test on 11 Italian sites, 56.8% of referring domains were spam.
Only 103 domains out of 3,365, 3.1%, were from the sites' own country.
Read domains, not links; strip out the noise before comparing.
Disavow is for manual actions, not for the red flags in a tool.
There's something the industry doesn't say much. A good share of the backlinks checkers count are produced by checkers themselves: on visilay.com, 430 referring domains out of 865 have "checker" in the name, and they're portals that publish automatic profiles of domains, each with a link. So the total number of backlinks is inflated, to a large extent, by the very tools that measure it. That's why in our reports the reference figure isn't the total but the relevant referring domains, which in today's sample, counting only .it domains, ranged from 0 to 20 per site. It's also the number you can genuinely grow, and it's the work we do in our SEO consultancy.
Until next time!
