Do your articles cite their primary sources in a way AI systems can follow?
Checks for machine-readable citation links so AI tools can verify the sources behind your claims.
What this signal tests
We parse the JSON-LD blocks on your article pages and look for the citation property on Article-class content. The citation can be a plain URL string or a richer CreativeWork object with a name and a URL. Each citation points at a primary source: a study, an interview transcript, a court filing, a press release, a previously published article. We confirm at least one such citation is present.
Why it matters for your visibility in AI
Modern AI answer systems use retrieval-augmented generation, which means they trust sources that themselves cite their sources. Articles with explicit citations are weighted more heavily in fact-grounding pipelines because the citation chain is auditable. An AI tool can follow your citation, verify what you said matches what the source said, and cite both. Without citations, your article is a dead end and is more likely to be quietly skipped in favour of a more transparent competitor. The consequence is visibility loss in exactly the AI surfaces that drive credibility traffic: scholarly Q&A, news summarisation, fact-check panels, and any vertical where readers expect provenance. For investigative reporting, scientific journalism, and policy analysis, missing citations are a major signal-quality penalty.
Pass criteria at a glance
| Criterion | Passes when |
|---|---|
| At least one citation on Article-class content. |
How we test it
We parse each Article block in your page's JSON-LD and look for a citation field. The field can be a single string URL, an array of URLs, a single nested CreativeWork object, or an array of CreativeWork objects. We accept all four shapes. We do not require the citation target to resolve, although we may flag broken citation links separately. The check completes during normal page parsing.
Show technical detection method
JSON-LD Article.citation present (string URL or nested CreativeWork); optionally verify target 200 over HTTPS.
If your site fails: how to fix it
- Identify the article templates that should carry citations. News, analysis, research, and explainer content are the typical candidates. Pure opinion or marketing pages do not need citation properties.
- Update each Article JSON-LD block to include a citation field. The simplest shape is an array of URLs: `"citation": ["https://example.com/study", "https://example.com/transcript"]`. The richer shape uses nested CreativeWork objects with name and url.
- If your CMS already records source URLs (WordPress with a custom field, Sanity with a references array, Contentful with linked references), wire those into the JSON-LD output. Otherwise, ask authors to paste source URLs into a citations field at publish time.
- Validate a sample page with Google's Rich Results Test or the Schema.org validator. Both will confirm the citation property is correctly typed.
- Re-run the AI Ready Test scan to confirm Article-class content carries citations.
Quick facts
| Maturity | ESTABLISHED |
|---|---|
| Weight | medium |
| Category | Trust & Provenance |
Primary sources
Related signals
Frequently asked questions
Will I need IT help to fix this?
Yes, but only for the template change. After the template is updated, authors can populate citations through whatever editorial workflow you already use. The developer cost is small and one-time.
Does adding citations affect SEO?
Positively but indirectly. Search engines do not rank specifically on citation count, but they reward content quality signals that citations correlate with: depth, originality, expertise. AI answer engines value citations more directly.
Should I cite every external link in my article?
No. Citation is for primary sources that ground specific claims, not for related reading or general links. A typical article has between one and ten citations. Over-citing dilutes the signal.
How long until the change takes effect?
On the next page render and recrawl. Once the JSON-LD is updated, AI crawlers and rich-results systems pick up the new citations on their next visit. Standard recrawl latencies (hours to days) apply.
Run your own scan
Run a free scan and see how your site grades across all 155 AI-readiness signals.