DomainScore is our own proprietary metric that shows the quality and strength of a domains backlink profile on a logarithmic scale of 0-100, higher is better.
You can get the DomainScore from our free bulk tool or 20 domains at a time using our simple API (try for free). And you can see the DomainScore in Rank compared to rank positions in Google and other metrics.
Compared to certain competing metrics we try to give a more authentic number simulating how we think Google might see the backlinks.
Examples of domains with a spammy link profile (updated 2019-03-07)
|Majestic TF / CF||Moz DA||SEOkicks|
|nikeolno.org||0||16||14 / 36||956||66%|
|christian-louboutin-sale.org||1||14||12 / 23||177||19%|
|nikerosheone.co.uk||0||25||15 / 29||2.945||81%|
|outletonline-michaelkors.com||0||32||16 / 39||2.693||98%|
|poloralphlaurenpascher.fr||0||22||13 / 26||1.565||48%|
|michaelkors-handbags.co.uk||0||30||14 / 25||2.019||77%|
|outletmichaelkors.co.uk||0||22||14 / 26||2.265||77%|
|fitflopsshoes.co.uk||0||32||14 / 25||2.185||78%|
|vitax.se||0||15||13 / 27||1.280||59%|
|rolace.co.uk||3||20||18 / 19||365||5%|
|moosehill.nu||0||5||6 / 22||404||18%|
|airmaxgeschaft.ch||6||49||21 / 22||72||19%|
|michaelkorshandbagsoutlet.co.uk||0||14||13 / 25||1.269||48%|
|pentlandpress.co.uk||11||48||23 / 19||228||33%|
Examples of domains with a non-spam link profile (updated 2019-03-07)
|Majestic TF / CF||Moz DA||SEOkicks|
|whitehouse.gov||100||91||89 / 76||85.668||100%|
|scstatehouse.gov||83||78||55 / 67||2.742||99%|
|momondo.com||69||74||33 / 47||1.920||99%|
|washingtonpost.com||100||92||86 / 91||197.622||100%|
|wordpress.tv||90||81||53 / 60||2.023||100%|
|wordpressfoundation.org||57||88||60 / 66||776||100%|
|skyscanner.se||50||53||30 / 60||330||86%|
|dus.com||93||77||49 / 61||1.401||98%|
|deutsche-bank.de||91||78||44 / 57||5.180||99%|
|diplomatie.gouv.fr||100||86||30 / 27||17.335||100%|
How we calculate DomainScore
We select up to 20 000 backlinks that matters from our own database pointing to the domain. This is only links from real websites ignoring spam, garbage, deep web, extra sitewide links and links from asia. In other words the cream of links.
Our formula looks at:
- The DomainScore of the domain that link (slight smaller version of DomainScore that is not recursive).
- If the backlink is on the home page or subpage and if the linking domain is a subdomain or not.
- The number of other links from the linking page, lower is better.
- The IP numbers of the domains that links. How is it spread, in what way? Natural or not?
- For which countries are the IP numbers hosted in, who is the owner (ISP) of the IPs, how is it distributed? Does it look natural?
- Is there any other domains that redirects to the domain and if so what is the DomainScore of those domains?
- The number of unique URLs and anchors used in all of the backlinks to the target domain.
- Very low quality links are excluded in the calculations.
Less Link Rating for subdomainsNovember 6th, 2018
Now when calculating DomainScore we decrease the link rating points from links that comes from subdomains.
Faster API with better limitsNovember 6th, 2018
With our simple API you can now get DomainScore of 20 domains at a time instead of 10 and it is now also much faster.
Reducing points for spamNovember 2nd, 2018
By not adding points from links with a very low link rating in the calculations I was able to reduce the DomainScore of domains with a spammy link profile without affecting domain names with a normal link profile. This resulted in for example this domain having DomainScore reduced from 12 to 0 that it deserves.
Link RatingNovember 1st, 2018
The link rating was improved and finalized. The output of this data was made public and live on the domain lookup pages ("Quality Distribution").
Faster performanceOctober 10th, 2018
With a Cassandra data remodel and some refactor of the code I was able to improve the performance resulting in a 95% speed increase in calculating DomainScore!
New Version of DomainScoreJuly 25th, 2018
Our new version of DomainScore is now done and live. Major changes includes that it now checks the domainscore of the domains that link and a new scale from 0-100 instead of 0-10. This will now be calculated and saved in the database for all the domains and be replacing the old domainscore across the system.
The DomainScore formula was programmed by Jim Westergren and it is the code that he is most proud of. It takes around 0.15 seconds to calculate the DomainScore of a domain with many backlinks (without cache) using our 9 dedicated servers (1 php, 1 mysql, 2 redis and 5 cassandra). We cache the number for 15 days.
Of course the DomainScore is going to differ a lot from other competing metrics but give it a chance, you might be surprised how good it is :)
Written by: Jim Westergren
Last updated: 2018-11-06