How to compare the links to your page and your competitors'

Most backlink gap reports compare whole domains. They tell you a competitor has 4,000 linking domains and you have 1,200, and leave you with a number too big to act on and too vague to be true for any one search.

Google ranks pages, not websites. The page that beats yours for a particular search usually has its own, much smaller set of links, and that set is what you're up against. A domain-wide gap mixes in links to a competitor's blog posts, careers page and press coverage, none of which help their product page rank for the search you care about.

This guide shows how to calculate the gap for one page and one search: which results to compare against, how to count referring domains page against page, how to find the domains that link to several competitors, what to leave out, and how to turn it into a number of links you can plan and price.

Start with one search and one page

A gap only means something against a specific results page. Pick the page you want to move and the one search that matters most to it: usually a buying or comparing search with real demand. If the page's main search is a learning search, links are usually the wrong tool (the guide to high-traffic keywords that deserve no link budget explains why).

Use your average position from Search Console over the last 28 days rather than one spot check. Rankings move by the day, by location and by device, and a single search from your own browser is a poor sample.

Choose the comparable results

List the pages ranking above yours. Then remove the ones you can't or shouldn't compete with on links:

• Marketplaces and large directories. A category page on a big review or app marketplace ranks on the strength of the whole site and its user reviews. You won't match that page by page, and you may be better off being listed on it.
• Wikipedia and reference sites. Their authority comes from the site as a whole and from how many people cite them. Nothing you buy closes that.
• Government and official pages. Often ranked for trust, not links, and not a fair comparison for a commercial page.
• Publisher roundups you'd rather appear in. A "best tools" article on a large publication is a page to get mentioned on, not one to outrank.

What's left are your comparable results: pages from businesses like yours, doing the same job as your page. Those are the pages you can realistically pass.

Then decide which of them you're aiming at. The target is the next realistic position, not number one. Moving from eighth to fifth is often achievable with a modest gap. Moving to first usually means competing with the strongest page in the market.

Count referring domains page against page

For each comparable page, and your own, pull the referring domains: the distinct websites that link to that exact URL. Count domains, not individual links. Fifty links from one site's footer are one referring domain and carry far less weight than fifty different sites.

This is the step that domain-level tools skip, and it changes the answer. A competitor's site might have ten times your linking domains while the page that outranks you has only double yours.

Record, for each competitor page, which referring domains also link to your page. Those are already covered. The ones that link to them and not to you are the gap.

Find domains that link to several competitors

A domain that links to one competitor might have done so for a reason unique to that company: a partnership, an investor, a customer story. A domain that links to two or three of the pages above you is telling you something different. It covers the topic and links out to the businesses in it. You're the one it's missing.

So sort the gap by how many competitor pages each domain links to. Domains that link to all of them come first, then those that link to two, then the rest. The first two groups are your shortlist: they are the most relevant and the most likely to consider linking to you as well.

Remove what you can't or shouldn't use

Before sizing anything, strike out domains that don't belong in the count:

• Competitor-owned or connected sites: their own blogs, sister companies, and integration partners that only list their own customers.
• Links you can't earn: news coverage of a competitor's funding round, or a supplier's case study about them.
• Low-quality or irrelevant sites: scraped copies, off-topic sites and pages that look built to sell links. Google's spam policies describe the link schemes it treats as spam, and a domain that looks like one is a risk, not an asset.
• Dead or redirected pages: a link report can list links that no longer exist. Spot-check the ones that matter.

A worked example

An invoicing software company wants to move its /invoicing/small-business page for "invoicing software for small business". It sits at position 7. All figures are illustrative.

Six pages rank above it:

Three comparable pages remain. Here are their referring domains, page against page:

Each gap figure is referring domains minus those shared with you: 62 - 6 = 56, 38 - 5 = 33, 27 - 4 = 23.

Adding those gives 112, but that double counts domains linking to more than one competitor. Sorting the gap by overlap:

Check it against the first table: A's gap is 7 + 9 + 5 + 35 = 56, B's is 7 + 9 + 4 + 13 = 33 and C's is 7 + 5 + 4 + 7 = 23. The 80 distinct domains are 112 minus the double counting (the 18 domains in exactly two lists are counted once too often, and the 7 in all three are counted twice too often: 112 - 18 - 14 = 80).

Now the exclusions:

The exclusions fall mostly on single-competitor domains, which is typical: that is where partner sites, funding coverage and low-quality directories collect.

Size the gap

The next realistic step for a page at 7 is position 5, held by Competitor B with 38 referring domains. Reaching parity with B means about 38 - 14 = 24 more referring domains. Passing C at position 6 needs about 27 - 14 = 13. Matching A at position 3 would need about 62 - 14 = 48, which happens to be every domain worth pursuing and is not a realistic first target.

So the gap to size is roughly 24 domains, and the shortlist writes itself: the 20 domains that link to two or three competitors (6 + 14), plus 4 from the single-competitor list, favouring those that link to B.

Parity is a gauge, not a promise. Twenty-four relevant domains that also link to your competitors are worth more than twenty-four random ones, and you may pass B with fewer. Plan the links in order of overlap and check your position as they land.

Two practical points before you price it:

• Not every domain is available. Some will only link after outreach, some never. If you can only close part of the gap, say so and size the expected gain on what you can close.
• Spread it out. Twenty-four new linking domains to one page in a single month looks nothing like how pages naturally attract links. Close a gap over several months. The guide to allocating an SEO budget across pages covers pacing.

This is the comparison Scale in Ship/Scale runs for each eligible page: it compares the page with the pages ranking above it for its main search, prioritises domains that link to several of them, and prices closing the gap from PlacementAPI's marketplace. Whether the price is worth paying is a separate question, answered in the guide to backlink purchases that can't pay for themselves.

Where this goes wrong

• Link counts are estimates. No link tool crawls the whole web, and Google doesn't publish the links it counts. Two tools can disagree about the same page by a wide margin. Use one source consistently so at least the comparison is fair, and treat the gap as a range.
• Relevance matters more than count. Ten links from sites your buyers read can outweigh thirty from general directories. The overlap sort helps, but look at the shortlist yourself.
• Content can matter more than links. If the pages above you answer the search better (clearer pricing, a comparison table, answers to the questions buyers ask), links may not move you at all. Compare the pages as a reader before you compare their links. The guide to improving existing pages or publishing new ones helps with that call.
• The results page changes. A competitor can gain links, a new page can arrive, and the excluded marketplace can drop out. Recalculate before each round of spending, not once a year.
• One search is not the whole page. A page ranks for many searches. The gap for its main search is the right one to plan against, but a page can gain from links in ways one search won't show.

The key points

Pick one page and its main buying or comparing search, list the pages above it, and drop the ones you can't or shouldn't compete with. Compare referring domains to those exact pages, not their whole sites, and sort the domains you're missing by how many competitors they link to. Strike out what you can't earn or shouldn't want, then size the gap against the next realistic position, not number one. Treat the result as an estimate, and check the content before you pay to close it.