Backlink CSV Analyzer
Summarises a backlink CSV: referring domains, anchor shares, rel split and top pages.
Root domains use a built in list of common two part suffixes such as co.uk and com.au, not the full Public Suffix List. Everything runs in your browser.
Result
0URLs read
0Unique domains
0Lines skipped
| Domain | URLs |
|---|
This free tool lets you extract domain from URL lists in one step: paste backlink URLs, get every unique root domain with the number of URLs it appears in, and copy or download the result as a plain list, a CSV or a disavow style list of domain: lines.
Quick answer
To extract the domain from a URL, take the host name, remove the scheme, port, path and www, then keep the part a person can register: example.com, or example.co.uk where the suffix has two parts. This tool does that for a whole list, counts URLs per domain and removes duplicates.
What does a referring domain extractor do?
A referring domain extractor turns a list of linking URLs into the list of websites behind them. Ten links from one blog are one referring domain, so counting domains instead of URLs gives a truer picture of how many sites link to you. Backlink tools and Google's Search Console Links report work the same way: Google says linking sites are "grouped by root domain", with protocol and subdomains removed.
The hard part is knowing where the registrable name starts. MDN explains that a registrable domain sits directly under an effective top level domain such as .com or .ac.uk, and that "it's impossible to tell algorithmically" which suffixes are public. The Public Suffix List records them. This tool uses a compact built in list of the most common two part suffixes, not that full list.
How do you use the Referring Domain Extractor?
- Paste your URLs, one per line. Lines from a spreadsheet work too: the first URL or domain on each line is used.
- Choose Root domain to merge subdomains, or Full host to keep blog.example.com apart from shop.example.com.
- Pick the sort order and the output: plain list, CSV with counts or disavow style domain: lines.
- Check the counters and the skipped lines list, so you know which lines were not read as URLs.
- Copy the output or download it as a .txt file, then review it before you use it anywhere.
How does the tool handle www, subdomains and ports?
It drops the scheme, any user name, the port and the path, removes a trailing dot and lowercases the host, then cuts it back to the root domain. So https://www.example.com/a, http://blog.example.com:8080/b and //cdn.example.com/c all become example.com. Lines without a scheme, such as example.org/page, are read as web addresses. IP addresses are kept as they are, because they have no root domain.
How are co.uk, com.au and other two part suffixes handled?
A built in list of 76 common two part suffixes, such as co.uk, org.uk, ac.uk, com.au, co.nz, co.in, com.br and com.pk, keeps one more label, so news.example.co.uk becomes example.co.uk. The Public Suffix List gives "com, co.uk and pvt.k12.ma.us" as examples of public suffixes and contains thousands of entries, so rare or three part suffixes are not covered here. Use Full host mode if a result looks too short.
Can I use the output as a disavow file?
You can use it as a starting point. Google's disavow help page asks for one URL or domain per line, a "domain:" prefix for domains, a .txt file "encoded in UTF-8 or 7-bit ASCII", and says lines that begin with # are ignored. It also warns the tool "can potentially harm your site's performance" if used wrongly, and says to disavow only when there are many spammy links that caused, or likely will cause, a manual action. Remove every domain you want to keep.
Quick reference: input and output
| Input | Root domain | Full host |
|---|---|---|
| https://www.example.com/page | example.com | example.com |
| http://blog.example.com:8080/a | example.com | blog.example.com |
| https://news.example.co.uk/ | example.co.uk | news.example.co.uk |
| https://myblog.blogspot.com/ | myblog.blogspot.com | myblog.blogspot.com |
| https://bücher.example/ | xn--bcher-kva.example | xn--bcher-kva.example |
| 192.0.2.10/page | 192.0.2.10 | 192.0.2.10 |
Worked example
Example
Example: the default list has 13 lines. 12 are read as URLs and 1 line, "not a url", is skipped. They reduce to 9 unique root domains. example.com appears 3 times (www, bare and a blog subdomain on port 8080) and example.co.uk twice. The other 7 appear once each: 192.0.2.10, another.blogspot.com, example.com.au, example.net (from the protocol relative //cdn.example.net), example.org, myblog.blogspot.com and xn--bcher-kva.example, which is bücher.example in Unicode. The two blogspot.com blogs stay separate because the platforms option is on.
Learn more
Frequently asked questions
What is the difference between a URL and a domain?
A URL is the full address of one page, such as https://blog.example.com/post?id=1. The domain is the name of the site, here example.com. Many URLs on one site share the same domain, which is why domain counts are smaller.
Why are blogspot.com blogs not merged into one domain?
Each blog on a hosted platform usually has a different owner, so treating myblog.blogspot.com as its own site is more useful in a link audit. Untick the platforms option if you prefer to merge them into blogspot.com.
What does xn-- mean in a domain?
It is the ASCII form, called Punycode, of an international domain name with letters outside A to Z. bücher.example becomes xn--bcher-kva.example. Tick the Unicode option to show the readable form instead.
Is this the full Public Suffix List?
No. It is a compact list of 76 common two part suffixes. Rare suffixes, private suffixes not in the platforms list and three part suffixes are not covered, so a few domains may come out one label too short or too long.
Are my URLs sent anywhere?
No. Everything runs in your browser with JavaScript, and the download is created on your device. Nothing is uploaded or stored, so you can paste client backlink lists safely.
How many URLs can I paste at once?
Tens of thousands of lines work in a modern browser. The table shows the first 100 domains and the output box always holds all of them, so copy or download from the box for large lists.
Sources
Checked 23 Oct 2026. Facts checked against these official pages. Read how I check facts in my editorial policy.