Backlink Leaderboard
FreeMalta Backlink Leaderboard · Methodology
Why DR alone
isn't enough.
A DR 94 site with 2.5 million outbound links gives you far less equity than a DR 72 site with 38,000. Domain Rating measures how strong a site is — not how much of that strength reaches you. We built a metric that does both.
LDS formula
680M+ AI citations analysed
Ahrefs · Semrush · Featured.com data
The problem
Domain Rating tells you one thing.
You need to know something else.
DR measures the strength of a site's own backlink profile. It says nothing about how many other sites that domain is pointing to — and equity isn't duplicated, it's divided. A page linking to 2,000 other sites passes 2,000× less equity per link than a page linking to one site.
SEO professionals have known this for years. 89% of experts flag spammy outbound links as the biggest red flag when assessing a site for link placement (Editorial Link, 2026). The platforms with the highest DR are often — paradoxically — the worst sources of link equity, because their authority is spread across millions of outbound links.
The FreeMalta leaderboard makes this explicit: every platform is scored on Link Dilution Score — a composite metric that accounts for DR, outbound dilution, dofollow status and verification. The result: platforms you've probably never heard of rank above Trustpilot and Forbes.
The metric
Link Dilution Score —
the formula.
Built from first principles and calibrated against published Ahrefs and Semrush link equity research. Every variable is explicit. Every calculation is deterministic.
Research basis
What the data says about
link dilution.
The LDS formula is grounded in published research from Ahrefs, Semrush, Backlinko and BuzzSumo. These numbers are not estimates — they are the published findings that shaped the formula constants.
01
Link dilution · Ahrefs 2024
73% more equity
Pages with fewer than 100 outbound links retain 73% more link equity than link farm pages with 300+ outbound links. This is the foundational evidence for the dilution component of LDS. A site with 38,000 outbound links is not 380× worse than a site with 100 — but it is meaningfully worse, and the power function in LDS captures this sublinearly.
Source: Ahrefs Link Equity Research, 2024
02
Outbound links · SEO expert survey
89% flag outbound spam
89% of SEO experts identify spammy outbound links as the biggest red flag when assessing a site for link placement. This is consistent with the dilution hypothesis: high outbound link count signals low editorial control, which Google's SpamBrain now detects with 94%+ accuracy (Moz, 2025).
Source: Editorial Link SEO Expert Survey, 518 respondents, 2026
03
Link longevity · Ahrefs 2026
71.2% link decay before 2020
An Ahrefs study tracking 2.3 billion backlinks over a decade found that 71.2% of links acquired before 2020 now return 404 errors or have been removed. Average lifespan on news and media sites: 18.4 months. This is why the leaderboard favours stable platforms — directories and review sites — over media placements. The LDS verification factor (1.05) rewards platforms with confirmed, persistent listings.
Source: Ahrefs Link Longevity Study, 2.3B backlinks tracked, 2026
04
DR vs ranking impact · Ahrefs
5.2× difference by DR tier
Links from high-authority domains (DR 70+) have 5.2× more ranking impact than links from low-authority sites (DR 30 or below). A single DR 80+ link can equal 10–20 links from DR 30–40 domains. This is the authority component of LDS — but it is bounded by dilution. A DR 80 site with 2M outbound links can score lower than a DR 60 site with 25,000.
Source: Ahrefs Ranking Factors Analysis, 2024
05
Featured.com original research
Brand-owned content cited 13× less
Featured.com's citation analysis found that third-party platforms — directories, review sites and communities — are cited by AI answer engines at 13× the rate of brand-owned content. Additionally, 34.5% of AI citations go to domains with DA under 40, and 76% of cited domains appeared exactly once across the dataset — indicating there is no static "citation shortlist" to get onto.
Source: Featured.com AI Citation Analysis, 2026
In practice
The same DR. Very different LDS.
Real examples from the leaderboard.
These six platforms illustrate why the leaderboard diverges from a simple DR sort.
Platform
DR
Est. outbound
Link type
LDS
Trustpilot
94
2,500,000
Nofollow
3.1
findly.tools
79
38,000
Dofollow
10.0
GitHub
96
12,000,000
Dofollow
9.3
Crunchbase
90
800,000
Nofollow
2.4
Twelve Tools
81
42,000
Dofollow
10.0
Medium
94
9,000,000
Nofollow
1.6
The key insight: GitHub (DR 96, 12M outbound links, dofollow) scores 9.3. Trustpilot (DR 94, 2.5M outbound links, nofollow) scores 3.1. findly.tools (DR 79, 38K outbound links, dofollow) scores 10.0. DR alone would rank them entirely differently.
The AI layer
680 million citations later —
what AI engines actually cite.
In 2026, an increasing share of information discovery happens through AI answer engines — not Google. The five major AI surfaces (ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews) have been analysed across 680+ million citations. The patterns are consistent — and they don't match traditional SEO logic.
11%
of domains cited by ChatGPT are also cited by Perplexity. Each engine has its own citation logic.
Averi, March 2026 · 680M citations
34.5%
of AI citations go to domains with DA under 40. Authority alone does not drive AI citation.
Featured.com citation analysis
9.9×
growth in AI referral traffic in 19 months. 92.4% from ChatGPT, converting at 7.1%.
Previsible, July 2026
76%
of cited domains appeared exactly once across all citation data — there is no shortlist.
Featured.com citation analysis
64%
of Claude citations go to brand-owned websites (SaaS/tech queries). Claude is different from every other engine.
OtterlyAI, 379K citations, June 2026
13×
more likely to be cited from third-party platforms (directories, review sites) than brand-owned content.
Featured.com citation analysis
Each engine cites differently.
One strategy doesn't cover all four.
The five major AI surfaces reach the same users with materially different citation behaviours. Being cited by one engine does not predict citation by another.
The 15 AI-cited platforms
Which platforms appear in
AI answers about your business.
These 15 platforms appear consistently across multiple independent citation datasets as sources that AI answer engines draw on when answering questions about companies, products and tools. A listing here feeds directly into how AI systems describe your business — regardless of dofollow status.
Sources
Research cited in this methodology.
| Source | Finding used | Dataset |
|---|
| Ahrefs Link Equity Research (2024) | 73% more equity from pages with <100 outbound links | Link equity analysis |
| Ahrefs Link Longevity Study (2026) | 71.2% of pre-2020 links now dead; 18.4mo media lifespan | 2.3B backlinks tracked |
| Ahrefs Ranking Factors (2024) | DR 70+ links have 5.2× more impact than DR 30 links | SERP correlation study |
| Profound / 5WPR (Aug 2024 – Jun 2025) | 680M citation baseline; platform-by-platform citation share680M citations |
| Attrifast (Apr–May 2026) | Citation density by engine: Perplexity 6.4, Claude 3.6, ChatGPT 3.1 | 14,400 prompt-runs |
| OtterlyAI (June 2026) | 64% of Claude citations go to brand-owned sites | 379K citations, SaaS/tech |
| Featured.com (2026) | 13× cited from third-party platforms; 34.5% citations to DA <40; 76% one-time cited domains | Citation analysis |
| Editorial Link SEO Expert Survey (2026) | 89% flag outbound spam; 64.1% prefer Ahrefs DR as authority metric | 518 respondents |
| Moz SpamBrain Analysis (2025) | Manipulative link detection accuracy: 94%+ | SpamBrain classification |
| Previsible (July 2026) | AI referral traffic grew 9.9× in 19 months; 7.1% conversion rate | Traffic analytics |
Questions
Methodology FAQ
Why does FreeMalta use LDS instead of Domain Rating?
Domain Rating measures a site's backlink authority — not how much of that authority reaches you. A DR 94 platform with 2.5 million outbound links distributes its equity across all of them. A DR 72 platform with 38,000 outbound links concentrates far more per link. The LDS formula quantifies that dilution explicitly, giving a 1–10 score that reflects the real equity transfer — not just the raw authority of the linking domain.
Where does the 50,000 constant in the LDS formula come from?
50,000 represents our calibrated threshold for "undiluted" outbound links on a typical curated platform. Pages with fewer than 100 outbound links retain 73% more equity than link farm pages with 300+ links (Ahrefs, 2024). We model this with a (50,000/O)^0.4 dilution factor — a power function that reflects the diminishing return of dilution reduction above the threshold. At 50,000 outbound links, you receive roughly the same equity as an undiluted page. Above that, the score declines.
What is the AI Score and how is it different from LDS?
The AI Score measures how likely a platform is to be cited by AI answer engines — ChatGPT, Perplexity, Claude and Google AI Overviews — when those engines answer questions about your company, product or industry. LDS measures link equity for traditional search rankings. The two metrics measure fundamentally different things and often diverge: Trustpilot has a low LDS (nofollow, high outbound dilution) but a high AI Score (actively cited for entity verification). Both are in the table.
Are the 15 AI-cited platforms based on original research?
They are based on a combination of: the FreeMalta leaderboard's observed citation data, the 2026 state-of-AI-citations synthesis (680M+ citations across ChatGPT, Claude, Gemini, Perplexity and AI Overviews from Profound/5WPR/Attrifast), and platform-level citation studies by Semrush, Ahrefs Brand Radar and OtterlyAI. Platforms marked AI cited appear across multiple independent datasets as consistent citation sources.
How often is the leaderboard updated?
The scoring formula is static and recalculates on every page load from the data.php dataset. The underlying platform data — DR, traffic, dofollow status, pricing — is reviewed quarterly. AI citation status is updated when a platform appears in or disappears from major citation studies. The formula itself will evolve as the relationship between outbound link count and equity transfer becomes better understood through published research.
