Link Analysis

A link analysis maps the authority your domain has earned through links, and sets it against the domains you compete with in the results. It is the diagnostic step that comes before any link work, because it answers the only question that matters at the start: how far behind or ahead are you, and on which pages.

Skipping it is how link budgets get spent on the wrong targets. Plenty of sites do not need more links at all. They need the authority they already have pointed at different pages, which you cannot see without looking properly. Everything below sits inside the wider discipline of linkbuilding, and inside SEO as a whole.

The numbers that actually tell you something

Referring domains is the metric to lead with. It counts unique domains linking to you rather than individual links, which makes it the closest thing available to a count of independent recommendations. One domain linking to you four hundred times through a sidebar is still one recommendation.

Domain Rating (DR) estimates the strength of a domain's link profile on a 1 to 100 scale. Two things about it get constantly misread. It is not a Google metric, so treat it as a comparative instrument and not a score to optimize. And the scale is logarithmic: getting from 59 to 60 is far harder than getting from 9 to 10, and a DR 20 domain is meaningfully more than twice as strong as a DR 10 one. The same idea exists at page level as URL Rating, based on the links pointing at one specific URL.

Anchor text distribution is where an unnatural profile shows itself fastest. A healthy distribution is mostly brand names and bare URLs, with descriptive anchors on the pages that earned them. A cluster of identical commercial anchors pointing at one landing page is the exact pattern the spam systems were built to find. Our guide to anchor text strategy covers how to keep it honest.

Topical relevance of the linking domains. A link from a site in your field is worth several from unrelated domains with bigger numbers attached. This is the part no tool scores for you, and the part worth spending human attention on.

Link velocity and lost links. The direction and pace of change over the last twelve months, plus the links you used to have. A steady climb reads very differently from two hundred domains arriving in a single month, and recovering broken or removed links is usually the cheapest work on the list.

Distribution across pages. How much of your authority lands on the homepage compared with the pages you actually need to rank. On most sites the answer is badly skewed towards the homepage, and rebalancing it through internal linkbuilding is faster than earning anything new.

How to run one

1. Pull your own profile. Referring domains, DR, anchor distribution, most-linked pages and the twelve-month trend. Ahrefs and Semrush both do this properly, and Bing Webmaster Tools gives you a usable free version, including for domains you do not own.
2. Pick a real competitor set. Not the companies you compete with commercially, but the domains holding the top positions on your priority keywords. Those are often two different lists, and the results page is the honest one. This work overlaps directly with a competitor analysis.
3. Compare like for like. Referring domains against referring domains, profile shape against profile shape. If your competitors sit at 400 referring domains and you are at 60, that gap is your brief.
4. Run a gap analysis. Find the domains linking to two or more competitors but not to you. Those sites have already demonstrated that they cover your subject and that they link out about it, which makes them the shortest route to a relevant link.
5. Audit your own quality. Look for the patterns that carry risk: links from domains with no traffic, footer and sidebar placements at scale, sponsored posts with no attribution, and networks of sites that link to each other and little else. Any link placed in exchange for money needs rel="sponsored", and an unattributed one is a liability rather than an asset.
6. Write the findings as a plan. Which pages need authority, which anchors are missing, which domains to pursue, and which existing links to clean up or reclaim.

Reading the numbers honestly

Volume is not the story. Artland went from a Domain Rating of 35 in May 2019 to 63 in April 2020, with links from Washingtonpost.com, Forbes.com, Observer.com and Nytimes.com. What moved the number was a small set of links from domains that were genuinely hard to get, placed in articles where the mention made sense to a reader.

The test we apply to any link, before and after acquiring it, is deliberately blunt: would you want this link if search engines did not exist? A link that sends real readers from a publication your buyers trust passes it. A link on a page nobody visits does not, whatever its DR says.

Treat DR as a symptom rather than a target while you are at it. It reports how link acquisition is going. It does not pay anyone's invoices, and a rising DR next to flat commercial rankings usually means the links are landing on the wrong pages.

What the analysis leads to

The output should be a short list, not a dashboard. Two or three pages that need authority, a handful of domains worth pursuing, the reclamation list, and anything that needs cleaning up. Then the work itself: earned coverage through digital PR, and internal links moving the authority you already have to the pages that convert.

A link analysis is part of every free SEO analysis we build, which is the quickest way to see where your profile stands against your market. When the finding is that you have a real gap to close, link strategy is the delivery that closes it.