how to run a backlink gap analysis for free

a backlink gap analysis answers one question: which sites link to my competitors but not to me in this dataset? compare referring-domain lists from a consistent snapshot, subtract your own, and rank the candidates. then inspect the linking pages, current links, audience and contribution policy before outreach. overlap is a research signal, not proof of editorial fit. our free backlink research guide helps you choose the source before following this checklist.

most link prospecting starts cold: you find a site, guess whether it takes external links and whether your topic fits. a gap analysis adds evidence of a recorded competitor relationship. the underlying link may be old, user-generated or on an irrelevant page, so qualification still matters. this guide covers set subtraction, free previews and the checks needed before drafting a pitch.

what a backlink gap analysis actually is

the mechanics are simple set math. take the referring domains of two or three competitors, take your own referring domains, and compute the set of domains that link to the competitors but not to you. that difference is the gap. each domain in it is a site that links into your category and has, so far, skipped you.

the relationship gives you a place to investigate. a recorded domain link does not establish its exact page, current status, topic or placement type. confirm those details before describing the site as a prospect.

what overlap tells you, and what it does not

overlap can prioritize research when you have relevant competitors and a useful dataset. it does not establish better reply rates, conversions or return on investment. measure those outcomes separately for your campaign.

there is a second-order benefit. domains that link to multiple competitors can be worth inspecting early. the page might be a relevant resource or directory, but it might also be a platform or an unrelated page. domain-level overlap alone cannot tell the difference.

the gap column is not magic. it does not work if you have no real competitors, or if your category is so new that nobody links to anyone. but for any established space, it is the first list you should build and the last one you should run out of.

keep the output as candidate rows, not guaranteed opportunities. the downloadablecompetitor backlink analysis template gives each row a place for its dataset, observed competitor relationship, current source-page evidence and qualification status, so an unverified domain does not become an outreach recommendation by accident.

the free step-by-step process

use available free data for a small manual shortlist. full exports and automated gap jobs may require paid access; record limits alongside the snapshot so an incomplete export is not mistaken for a complete profile.

step 1: pick 2 to 3 real competitors

not aspirational competitors. real ones, ranking for the same queries you want, at roughly your size or a step above. pick two or three, not ten. the overlap math gets noisier with more competitors, and three is enough to surface the domains that link across the whole category. choose sites whose audience genuinely overlaps with yours, not just any big name in the industry.

step 2: pull their referring domains

for each competitor, get the list of unique referring domains. you cannot do this in google search console or bing webmaster tools, because those only show properties you have verified. you need a tool that indexes the whole web. pull your own referring domains the same way so you have all four lists (or three competitors plus yourself) in hand.

step 3: compute the overlap that excludes you

this is the set difference. take the union of the competitor referring domains, then subtract every domain that already links to you. what is left is the raw gap. if you are doing this by hand, drop each list into a spreadsheet column, dedupe, and use a lookup to flag the domains that appear for a competitor but never for you.

step 4: filter platform and cdn noise

the raw gap is dirty. it is full of domains nobody can realistically pitch or that are not editorial links at all. strip these before you rank anything:

• platforms and ugc: youtube.com, facebook.com, twitter.com, linkedin.com, medium.com
• cdns and asset hosts: cloudfront.net, akamai, gstatic.com, image and font hosts
• shorteners and trackers: bit.ly, utm-laden redirect domains, analytics endpoints
• your competitors themselves and obvious aggregator scrapers

these show up for almost every site, so they tell you nothing about the category by themselves. filter for your use case and report how many rows you removed; no fixed percentage applies to every dataset.

step 5: rank by authority

rank candidates by competitor overlap, then use authority as one secondary signal. inspect the linking pages and topical fit before deciding whom to contact. a high score alone does not establish quality or suitability.

step 6: pitch

first find the actual linking page and confirm the link is current. check audience fit and contribution policy. if contact is appropriate, reference the specific page honestly and explain how your evidence helps its readers. a domain-only result is insufficient for that claim.

doing the whole thing free

the bottleneck in the manual version is step 2: you cannot pull a competitor's referring domains from gsc. third-party tools have different preview, export and plan limits. if you are deciding which parts of that subscription you still need, the ahrefs alternative comparison by job separates owned-site audits, rank tracking, historical links, and gap analysis. crawlgraph runs the gap analysis for free on the common crawl webgraph (3.90B domain-level links across 121.1M domains, as of the current release), and it ships a dedicated gap-analysis plus outreach-target finder that does steps 3 through 5 for you: it computes the overlap, filters the platform and cdn noise automatically, and ranks the survivors by cg_authority (a 0 to 100 score). the free tier runs the whole gap workflow and shows your top 5 gaps right in the browser at the gap-analysis tool after an email step. paid access expands available results, csv export and api quotas. the lifetime option is $99 paid once; eligible visitors may also see a monthly option.

with paid api access, submit your domain and competitor domains to create a job, then poll its status. the following illustrative request and result show the shape; see the api reference for the current contract:

curl -X POST https://crawlgraph.com/api/v1/gap-analysis \
  -H "Authorization: Bearer cg_live_…" \
  -H "Content-Type: application/json" \
  -d '{
    "my_domain": "yoursite.com",
    "competitor_domains": ["rival-a.com", "rival-b.com", "rival-c.com"]
  }'

the result carries the gap domains with their authority score and the list of competitors each one was found on, so you can sort straight into an outreach sheet:

{
  "job_id": "gap_8c2f…",
  "status": "completed",
  "result": {
    "total_gaps": 1284,
    "gaps": [
      { "linking_domain": "industry-roundup.io", "cg_authority": 71, "found_on": ["rival-a.com", "rival-b.com"] },
      { "linking_domain": "niche-review.com",     "cg_authority": 58, "found_on": ["rival-a.com", "rival-c.com"] }
    ]
  }
}

paid api access includes 50 gap jobs and 1,000 backlink calls per utc month. free keys allow 15 backlink calls and no gap jobs. the same job is available to ai clients through the hosted mcp server at crawlgraph.com/mcp (or npx -y crawlgraph-mcp locally for claude, cursor, and cline), so you can ask for a gap list in plain language. full reference lives at /docs/api.

common mistakes to avoid

• too many competitors. ten competitors produces a huge, shallow overlap. two or three close rivals gives a sharper, more actionable list.
• ranking before filtering. if you sort by authority before stripping platform and cdn noise, the top of your list is youtube and cloudfront. filter first, rank second.
• counting backlinks instead of domains. dedupe to the referring domain. one outreach email goes to one site, no matter how many links it gave a competitor.
• ignoring the multi-competitor signal. the domains that link to several rivals are your best prospects. do not let them get lost in an authority-only sort.
• expecting same-week data. a link a competitor earned last week will not appear until the next common crawl snapshot lands. gap targets are stable, so this rarely bites, but it is worth knowing.

conclusion

a backlink gap analysis creates a research shortlist. pick two or three real competitors, pull their referring domains, subtract your own, strip the platform and cdn noise, rank what is left by overlap and authority, then inspect the actual linking pages. manual research can use free exports where available; crawlgraph's free gap preview shows five candidates after an email step. fuller results and exports require paid access.

the work is finding the overlap; the value is in pitching it well. start with the gap-analysis preview, see what the index has, and build your first gap list from there. for the data background, read common crawl, explained for seos. inspect a retained example in the seo-tool overlap study and the broader free backlink guide.