Backlink Analysis

A backlink analysis is a snapshot of every external reference pointing to a website at a single point in time: which domains link, out of which environment, with which anchor text, and how all of that compares to the competition. It answers the question of how a link profile is set up.

Not to be confused with monitoring. The analysis gives you one picture, monitoring gives you a series of them. If you want to know what changes and when you need to react, that is covered in the Backlink monitoring guide. This article is about recording the current state.

What Data You Need

Before the evaluation starts, it has to be clear where the numbers come from and what they cover.

Where the Data Comes From

• Google Search Console shows the links Google itself knows about. It is free and the only source that comes straight from the search engine operator, but it offers no quality assessment.
• Backlink tools deliver the inventory along with their own metrics. Many treat Ahrefs as the reference point for the size of the live index, Semrush covers competitor analysis and auditing broadly, SISTRIX is particularly widespread in German-speaking markets, and Seobility is a cheaper alternative for smaller projects. Each one runs its own crawler, which is why the numbers differ from one another.
• Your own campaign documentation shows which placements were commissioned and whether they actually went live.

The entirety of all incoming references is called the link profile. It is what every analysis looks at, and it consists of more than a single number: the referring domains, how they are distributed, the anchor texts, where the links sit and which attributes they carry.

What You Derive From It

Comparing these three sources is what produces the actual result: which links Google knows, which ones only the tools know, and which commissioned placements never arrived. That difference is exactly the part a pure tool export will not show you.

One note that saves a lot of disappointment: no tool sees every link. The deviation between two providers regularly lands in the double-digit percentage range, because they crawl different amounts of the web. Absolute numbers are therefore of little use, while ratios and comparisons are worth a great deal.

The Analysis in Six Steps

• Define the target pages the profile is actually relevant for.
• Export the inventory from at least two sources.
• Remove duplicates and obvious spam.
• Build metrics instead of judging individual links.
• Compare against the domains that are visible for your keywords.
• Turn the gaps into a list of priorities.

Why Step 1 Comes First

Analyzing the profile as a whole sounds thorough, but it rarely leads to an action. It only gets interesting per target page: a money page without external references is a finding, a blog overview without references is normal. In practice we see teams evaluate profiles for weeks and still not know afterwards which page needs support next.

Why Step 3 Cannot Be Skipped

Exports regularly contain the same domain several times over, plus directories, scrapers and automatically generated pages nobody ever commissioned. Count those rows in and what you are mainly measuring is the background noise of the web.

Metrics and What They Tell You

What Trust Flow and Citation Flow actually measure, and how to read the relationship between them, is covered in Trust Flow explained and Citation Flow explained. Domain Rating and Spam Score work on the same principle: useful estimates for pre-sorting, not verdicts.

Distribution Matters More Than the Total

A profile with 200 references from 180 different domains looks completely different from one with 200 references from twelve. The spread shows whether external authority comes from many independent sources or rests on just a few places. A broadly distributed profile typically comes out of visibility that grew over time, a narrow one out of few placements that were repeated.

Where the Link Sits on the Page

References inside the body text of an article are weighted considerably more heavily than links in footers, sidebars or link lists. The reason is straightforward: whatever sits in the navigation appears on every subpage and is therefore not a statement about any particular piece of content. When you evaluate the inventory, it pays to separate it by placement instead of counting every row the same.

Why Volume Is the Wrong Guiding Metric

Quality beats quantity, and in an analysis that shows up particularly clearly. Ten topically fitting references from publications with their own editorial team say more about a domain than a thousand entries from directories. Not every page needs backlinks, but pages with business value and competition need the right ones.

Competitor Analysis and Link Gap

The comparison is the part with the biggest practical payoff, because it replaces absolute values with relative ones. The question is not "are 400 domains a lot", it is "who links to the pages that rank ahead of me for my keyword".

Who You Should Actually Compare Against

A common mistake: the comparison gets made against the rival that sales talks about. For this analysis, though, what counts is who is visible for the relevant search queries. Those are often trade publications, comparison portals or knowledge sites, not the company one town over.

In our experience this is where analyses most often come to nothing: the result describes a set of competitors that compete for the same customers, but not for the same search results.

Making the Gap Visible

A link gap shows domains that point to several comparison sites but not to you. It is the most concrete list an analysis can produce, because every row has already proven that it writes about the topic.

So that this list does not have to be assembled by hand, MentionIQ analyzes your website, your competitors and your important keywords, and identifies suitable link sources and placements from that.

The Gap No Backlink Tool Shows

Backlink tools capture references. They do not capture mentions of a brand without a link — and for AI search systems those mentions are exactly what matters, because there it is the text an answer gets built from that counts. A competitor analysis that only looks at links therefore misses part of the visibility picture. How to measure that is covered in AI visibility.

What a Suspicious Profile Reveals

Suspicious does not mean guilty. It means: look closer.

• Many references from domains with no recognizable editorial team or audience of their own.
• Outbound links dominate the article inventory of the linking site.
• One anchor text repeats almost word for word across many domains.
• Growth arrives in bursts instead of continuously.
• Off-topic sources with no connection to the subject matter.

Paid placements explicitly do not belong on this list. An invoice does not make a reference suspicious; what is suspicious are patterns without any editorial connection. A sponsored post in a trade magazine, topically fitting and labeled as such, is a different thing from a hundred placements out of one network.

What You Take Away From the Analysis

An analysis that ends in a number was an exercise in diligence. It has to lead to three statements:

• Which target pages need external authority first?
• Which domains from the link gap are realistic options?
• Which existing references are so valuable that losing them would have to be noticed?

The third point is the handover to monitoring.

For the roles inside a company, the result means different things. SEO teams get a work list. Marketing leads get a justification for a budget, because the gap to the competition is quantified. For management, it becomes visible how far behind the site is and over what period that can be closed.

So that the choice of publishers afterwards does not come down to gut feeling: GetMentioned is a placement marketplace for backlinks, brand mentions and digital PR, where you select suitable publishers, compare SEO data and prices, and book placements directly. The GetMentioned Score brings together price, SEO data, authority and possible impact instead of judging sources by a single metric.

Conclusion

A backlink analysis is not a counting exercise, it is an aid to prioritization. Absolute numbers are unreliable because no tool sees the whole web. What makes the evaluation solid is ratios, the comparison against the domains that are actually visible, and the question of which target page needs support next. Everything else is a report nobody reads.