We Analyzed 7,000 Competitor Backlinks. Two-Thirds Were Useless.

Every "steal your competitors' backlinks" guide starts the same way: export their backlink profile, sort by authority, work down the list. We build a product in this space, so we decided to actually measure what that advice is worth. We pulled roughly 7,000 raw backlink rows across three real sites — a SaaS form builder, a privacy-first analytics tool, and a coffee content site — and checked, by hand, how many of those links a small site could realistically copy.

The answer: not many. In the worst case, 66% of all raw rows were structurally uncopyable before we even started judging quality. But buried in the remainder was a pattern that repeats across every niche we looked at — and once you see it, you can filter any competitor's backlink export down to a workable action list in an afternoon.

This post is the full breakdown: what the raw data actually contains, the three categories of links you can never copy, the three categories you almost always can, and the checklist we now run on every report.

The setup

Three sites, three niches, three confirmed competitors each:

Tally (form builder) vs. fillout.com, forms.app, paperform.co — the full write-up is in the Tally sample report
Fathom Analytics (privacy-first analytics) vs. plausible.io, simpleanalytics.com, umami.is — Fathom sample report
Homegrounds (coffee guides and gear reviews) vs. cornercoffeestore.com, coffeeness.de, roastycoffee.com — Homegrounds sample report

For each project we pulled competitor backlink rows from a licensed backlink index — the same class of data the big SEO suites resell — then fetched every shortlisted source page live and judged it by hand: is there a real page here, is the competitor actually linked on it, and could a comparable site get the same link?

The funnel for the Tally project looked like this: 7,000 raw rows → 4,621 discarded immediately (more on why below) → 97 candidate source domains → 23 shortlisted and verified live → 15 qualified opportunities. Homegrounds was even starker: 775 candidate domains after filtering, 19 verified, 7 qualified. Fathom, researched entirely by hand, went 45 candidates → 22 qualified.

Two-thirds of the "opportunities" evaporate under a single question: could we actually get this link? Here's where they go.

The three categories you can't copy

1. High-authority editorial coverage

The top of every backlink export — sorted by Domain Rating, the way every tutorial tells you to sort it — is press. TechCrunch covered their funding round. A major outlet quoted the founder. A big media brand included them in a review written by a staff editor with no submission process.

You cannot pitch your way into replicating these. The link exists because of a news event, a relationship, or a PR firm. When we ran automated competitor discovery for the coffee site, it dutifully returned tastingtable.com and seriouseats.com as top "competitors" — Dotdash and Condé brands with full-time editorial staffs. A niche content site replicating their link profile is not a strategy; it's a fantasy the raw export sells you.

The tell: the linking page is a news story or staff-written feature, and the only contact path is a generic tips inbox.

2. Customer-content pages

This was the single biggest bucket in the entire study, and almost nobody talks about it. For the form-builder project, 4,621 of 7,000 rows — 66% of everything — pointed at customer-form subdomains. URLs like forms.fillout.com/t/abc123: a link to someone's registration form, hosted on the competitor's platform.

These show up in the export as "backlinks to fillout.com." They are nothing of the kind. They're links to a customer's content that happens to live on the competitor's domain. There is no page to get listed on, no editor to email, no version of this link you can earn. Any tool that hands you a raw count without stripping these is inflating the opportunity by 2–3x.

The same family includes URL shorteners and email wrappers (tinyurl, Outlook SafeLinks, ConstantContact mirrors) and third-party support-desk instances — redirect junk and private docs that leaked into the index.

3. Paid placements masquerading as editorial

The third bucket is the sneakiest, because the pages look exactly like the replicable ones. During the Tally research we noticed a competitor's link on a "Typeform alternatives" listicle carried a utm_source=f-partners tag — a commercial partnership placement, not an editorial mention. A meaningful share of the competitor's replicable-looking links carried the same marker.

Other specimens from the reject pile: a press-release wire service whose "pickup" consisted of nofollow links on mirror sites (one already 404), and a once-legit SaaS directory that is now a parked domain for sale. If you copied your competitor's exact playbook here, you'd be copying their spend, not their SEO.

The tell: UTM parameters on the link, a "become a partner" page one click away, or every outbound product link on the page pointing at the same affiliate network.

The three categories you almost always can copy

Here's the good news. After the carnage above, what survived wasn't random — it clustered into three repeatable shapes, across all three niches.

1. Directories and listings with an open submission process

The highest hit-rate category in the study. If a directory has a public submit page and a competitor is listed while you're absent, that's a near-mechanical win:

aipure.ai — forms.app has a full product page with a dofollow link; the Tally slug 404s; the submit page promises review within two business days
GoodFirms — forms.app has a reviewed profile; searching "tally" returns only unrelated accounting software; the "Get Listed" flow is self-serve
awesome-selfhosted (GitHub, 312k stars) — Plausible and Umami are listed in the Analytics section; Fathom Lite qualifies under the published contribution rules; the "contact path" is a pull request
AlternativeTo — competitors cross-listed as alternatives on four high-authority pages where Fathom is absent; suggesting an alternative is a free account action

Several of these need zero outreach at all — the Fathom report flagged 7 of its 22 opportunities as pure self-serve wins you could finish in a week.

2. Listicles and reviews with a reachable editor

Mid-tier "best X tools" articles are the workhorse of competitor link profiles, and the replicable ones share a fingerprint: the article is actively maintained, and a human's email address is findable on the site. Real examples from the reports:

flippingbook.com links forms.app contextually in a SaaS customer-support article that was updated this year — an actively-maintained page with a verified contact page
icons8.com/blog holds a contextual dofollow link to forms.app — and runs a public become-an-author program with a named editor email
seline.com featured Fathom at #3 in a Google Analytics alternatives roundup with no outbound link, while competitors got linked — the founder's email is in the site footer. The fix is a thank-you email, not a pitch
studyfinds.com cites a coffeeness.de editor as the expert source in a coffee-gear listicle — a fixed "expert quote" slot that already exists on the page layout

That last one generalizes into the single most interesting pattern we found in content niches: dozens of mid-tier listicles contain a templated expert-recommendation slot ("This product was recommended by Kelsey Waddell from Roasty Coffee" appears verbatim across at least five near-identical articles). If a competitor's name is sitting in that slot, the pitch isn't "link to me" — it's "here's a named editor and a quotable take for your next update."

3. Podcast guest intake

Every niche has podcasts, and a surprising number have formal guest submission pages. Competitors' founders appear as episode backlinks, and the door is open:

Code Story ran the Simple Analytics founder's creation story; the guest-intake form is live
The Changelog gave the Plausible co-founders a full episode; it has a formal podcast-request form
Sustain and Data Diva Talks Privacy both hosted competitor founders, both with verified contact forms

A podcast episode is a durable link from the episode page, show notes, and usually a few aggregator mirrors — for the price of an hour of talking about your own product.

The checklist: filter any export yourself

This is the exact sieve we now apply to every report. Run your competitor's backlink export through these questions, in order, and stop at the first "no":

Is the link target actually the product? If the URL is a customer subdomain, shortener, or email wrapper, discard. (This one step killed 66% of our biggest dataset.)
Does the source page still exist and still show the link? Fetch it. We rejected dead mirrors, parked domains, and pages whose evidence couldn't be reproduced — assume nothing.
Is it press or a partnership? News coverage, investor pages, and anything with partner/affiliate UTM tags goes in the "can't copy" pile.
Is there an open door? A submit page, a write-for-us page, a guest-intake form, a PR process. If yes, it goes straight to the top of the list.
Is there a named human? An editor email, an author page, a founder in the footer. A generic contact form works; no contact path at all disqualifies the page no matter how good it looks.
Is the page alive? An article updated within the last year can be pitched an addition. One untouched since 2021 usually can't.

Expect the funnel to be brutal — ours ran from 7,000 rows to 15 qualified opportunities, and from 775 candidates to 7 in a content niche. That feels like failure until you realize the surviving 7–15 items each come with a verified page, a real contact, and an obvious next action. Fifteen links you can actually get beat 7,000 you can't.

Do it yourself, or let us do it

You can absolutely run this process by hand — everything above is the recipe. Pick your three closest search competitors, pull their backlinks from any index you have access to, and apply the six questions. Budget a weekend.

If you'd rather skip to the action list: replinks.co does this as a one-time report. Enter your domain on the homepage and the free analyzer returns 3 preview opportunities immediately; the full $29 report goes deeper — 20–30 qualified opportunities, each with the link evidence, the verified contact path, an effort and risk assessment, and an editable outreach draft. The three sample reports linked above are the exact format, warts and all, if you want to judge the output before spending anything.

Either way: stop sorting competitor backlinks by Domain Rating. Sort them by whether you can get them.