When a competitor publishes less than you and still ranks above you on the same keyword, one of the reasons is usually the sites that link to it. A link gap analysis reads that list in reverse: it finds the domains that link to your competitors but not to you.
The output is an outreach list. Below we cover how to build it, why you should not use it raw, and how to rank the rows.
What a link gap measures
The calculation is simple. You pull the referring domains for each competitor, compare them with your own list and keep the ones you are missing. A domain that links to two or more competitors moves up. That site is used to linking on this topic and already knows your industry.
The reason to work at domain level rather than page level is noise. One news site may have sent forty links to a competitor. For outreach there is still only one target: that site's editor.
Pick the right competitors
The analysis is only as good as the competitors you choose. Your business competitor and your search competitor are not always the same company. A content or comparison site you keep seeing on page one for your keywords may never show up in a sales call, yet it can be the most instructive competitor for its link profile.
Three to five competitors are enough for most work. More of them make the list longer but lower the share of domains that link to several competitors, and that share is the most useful signal in the list.
Why you should not use the raw list
A few kinds of rows dominate the first output. Automated directories and scraper sites link to every domain; they probably already link to you, or a link from them is worth nothing. The competitor's own network ends up in the list too: sister brands, sites from the same group, partner pages. There is no point reaching out to them. Paid placements and sponsored content should be flagged separately.
The rule: clean the list before you prioritise it. If you don't, the top is taken by the sites that link most, which are the least selective ones.
Prioritising the list
For each row in the cleaned list, ask:
- How many competitors does it link to? Two or more suggests the site covers this topic regularly.
- Which page does the link come from? A guide or comparison article can be updated; an old news archive usually is not.
- Does the site itself get traffic? A link from a page nobody reads brings no visitors either.
- Is the link relevant to your target page? A site that would link to your category page expects a different offer from one that would link to a blog post.
- Is there someone you can reach? A site with no contact route can stay on the list, but it moves down.
Outreach: what are you offering
A link gap tells you who links, not why. Open the page that links to the competitor and read the sentence the link sits in. Was it given for a piece of data, for a tool, or as one option in a comparison table? Your offer should match that. If you have fresher data or an option the table is missing, say so in the first sentence.
Paying for links is a separate matter. According to the spam policies page on Google Search Central, buying or selling links for ranking purposes, whether for money, goods or services, counts as link spam. The same page says paid links qualified with rel="sponsored" or rel="nofollow" do not violate the policy. Drawing that line when you build the outreach list is easier than cleaning up later.
Log every outreach in the list: who was contacted, what was offered, whether they replied, whether a link was added. After a few months your own data shows which kind of offer works. Without that log, two people write to the same site separately, or a rejected offer gets repeated word for word three months later. When a link is added, check on the next read whether the provider sees it too; if not, the problem may be the scope of the read rather than the link.
How often to repeat it
Link profiles change slowly. A monthly read is enough for most teams; a weekly one tends to reproduce the same list. What you really want to watch is the change. The new referring domains a competitor gained this month show which campaign or content earned them links.
Lost links are opportunities too. If a page that linked to a competitor was removed or its link broke, the owner may be looking for a new source. One measurement caveat: if the provider could not read a competitor's list in a given month, that month's new link count is unknown, not zero. We cover that distinction in detail in our post on tracking competitor social media.
With Adversee
Adversee's SEO module reads the referring domains of the competitors you track every month. It lists domains that link to at least two competitors, do not link to you and do not have a high spam score as link candidates. Since it reads the first 100 referring domains per competitor, it says so when a competitor's list was cut off instead of assuming it saw everything.
Start tracking competitor moves automatically.