Domain Credit Scoring
A domain can have a strong authority score and never once appear in an AI answer about your category. The reverse happens too. Domain Credit Scoring ranks outreach targets on the second thing, because that is the one that decides whether an engine quotes you.
Part of Research & Credit ScoringResearch
The challenge brands face
Authority scores were invented to approximate trust in a link graph. AI engines do not read a link graph when they answer — they pull from a set of sources they already treat as reliable for that topic, and that set is narrow. In our July 2026 study of 1,000 Indonesian queries, 930 returned an AI Overview and their 7,880 citations landed on just 1,578 domains, with the same names recurring answer after answer.
Knowing the list exists does not make it usable, which is where most link-building advice stops. The domains at the very top are platforms and category giants — in our data YouTube appeared in 58% of answers, and in health Alodokter and Halodoc each exceeded 81% — and no outreach email is getting you onto them. The long tail below is mostly one-off citations. The workable targets sit in the middle, and finding that middle takes measurement, not a bigger prospect list.
Why this matters now
Brand mentions on trusted pages now carry more weight for AI visibility than bare links do — Ahrefs' analysis of 75,000 brands found branded web mentions correlate with AI Overview visibility at 0.664 against 0.218 for backlinks. That changes what outreach should be pitching, and it makes the choice of domain more consequential than the choice of anchor text. The citation map also drifts as models update, so a target list built last year is describing an answer set that no longer exists.
Research
Our approach
- 1
Capture the citations, all of them
We run a category query set through AI answers and record every cited source — expanding the full source list rather than the handful shown by default, because stopping at the collapsed view undercounts citations badly. Each capture stores the title, source, snippet, URL, and domain, and the query set is written the way people actually type rather than in marketing language.
- 2
Measure citation rate per category, then profile it
Citation rate is the share of answers in a category that cite a domain — a domain cited in 211 of 257 health answers sits at 82.1%. We calculate it per category rather than globally, because an engine's trust in a domain is topical and mixing categories hides that. Every domain then gets a semantic profile built from the queries it was cited for, so two sites with identical rates can still be told apart by what they are trusted for.
- 3
Score, filter, and rank what you can reach
Candidates are ranked with a weighted score: 70% semantic similarity between your target query and the domain's citation profile, 20% normalized citation rate, and 10% citation position. Before scoring, every domain above the 95th percentile of citation rate in its category is dropped — those are the platforms that dominate answers and will never link out. What remains is a middle tier that gets cited consistently and still accepts contributions, quotes, and partnerships.
Research
What you receive
- A citation rate for every domain cited in your category, measured from live AI answers rather than a link index
- A semantic profile per domain showing which queries each one is trusted for
- A weighted opportunity score — 70% semantic similarity, 20% citation rate, 10% citation position
- A reachability filter that removes the top 5% of mega-domains you could never earn a link from
- A ranked outreach shortlist with the play to run on each domain: contributed expertise, a quote, a data contribution, or a roundup placement
Research
How we prove progress
Every engagement is measured against numbers you can check yourself.
- Citation rate per domain, per category
- Semantic similarity to your target queries
- Average citation position in source lists
- A ranked, reachable outreach shortlist
Research
Who this is designed for
Teams still buying links on Domain Authority while AI answers cite a different set of sites
PR and comms teams that need to prove which placements are worth pitching
Brands invisible in AI answers even though their own pages are in good shape
Agencies who need a defensible reason for every domain on the outreach list
Research
What to expect
Outcome
Your PR and link building aim at the domains AI engines already quote in your category, instead of the ones a link index rates highly.
An honest timeline
The citation map for one category takes about a week to build, and it is worth refreshing every quarter — the set of domains AI engines cite shifts as models update.
Research
Questions: Domain Credit Scoring
What is Domain Credit Scoring?
Domain Credit Scoring ranks backlink, PR, and citation targets by evidence that AI engines quote them in your category. Each domain is scored on three weighted components — 70% semantic similarity to your target queries, 20% citation rate, 10% citation position — after the top 5% of unreachable mega-domains are filtered out. The output is a ranked outreach shortlist, not an authority table.
What is a domain citation rate and how is it calculated?
Citation rate is the percentage of AI answers in a category that cite a given domain. Count each domain once per query it appears in, divide by the number of queries in that category, and multiply by 100 — a domain cited in 211 of 257 health answers has an 82.1% citation rate. It is calculated per category because AI engines trust domains topically, not universally.
Why do you exclude the biggest domains from the shortlist?
Because they will never link to you, so keeping them wastes outreach effort and crowds out realistic targets. Every domain above the 95th percentile of citation rate in its category is filtered out before ranking — platforms like YouTube, which appeared in 58% of answers in our study, or dominant health portals. Removing them keeps the ranked list actionable rather than aspirational.
Do backlinks still matter for AI search in 2026?
They help, but brand mentions matter more. Ahrefs' study of 75,000 brands found branded web mentions correlate with AI Overview visibility at 0.664 versus 0.218 for backlinks — roughly three to one. The practical move is to pursue placements that name and describe your brand on pages AI engines already cite, with or without a link, which is why the shortlist comes with a suggested play per domain rather than just a URL.
How often does the citation map need refreshing?
Roughly every quarter. The set of domains AI engines cite shifts as models update and as pages get refreshed, so a one-off scrape goes stale — a domain can lose ground in AI answers without changing anything on its own site. Freshness matters on the other side too: recently updated pages tend to earn more citations than pages left untouched for years.
Other services in this practice
- Keyword, Topic & Query ResearchFind the exact words buyers type into Google and the questions they ask ChatGPT
- Creator Credit ScoringScore creators on the health of their community, not the size of their following
Tell us what you need
Describe the problem, the team and where it has to run. An engineer reads every enquiry and replies by email.
30-minute call
Book a 30-min callEmail
[email protected]
You will get an engineer on the first call, not a qualification script.