Find the page that should be winning each prompt — or confirm it doesn't exist.
For every prompt that comes back not cited, Content Library Matching scores every page in your indexed content library against it using seven signals. The match outcome — strong match, borderline, or no match — determines the gap diagnosis path and the fix type your team acts on.
Available on Pro + Diagnose and above
The 7-signal hybrid matcher.
Content Library Matching runs a weighted scoring pass across every indexed page in your library for each unresolved prompt. Seven signals contribute to the match score — exact keyword match carries more than audience fit overlap.
Exact Keyword Match
Does the page's primary keyword match the prompt's primary keyword?
Close Keyword / Synonym Match
Are the keywords related via known aliases? (e.g. "session replay" ↔ "session recording")
Same Topic Cluster
Do the page and prompt share a normalized topic cluster?
Same Intent Served
Do the page and prompt serve the same intent type?
Title Token Overlap
How much do the page title and prompt query share when tokenized?
Key Entity Overlap
Do the page and prompt share key entities — brands, products, concepts?
Audience Fit Overlap
Do the page and prompt target the same audience roles and personas?
Three match outcomes.
Score ≥ 65
Strong match
The highest-scoring page is the candidate. Gap diagnosis runs immediately: your matched page vs. the competitor page currently being cited for this prompt.
Score 40–64
Borderline
Up to three candidate pages go to an AI tiebreaker. The tiebreaker also runs a cannibalization check: if two of your pages are both strong candidates, the matcher selects the more specific one and flags the other for review.
Score < 40
No match
No matching page exists in your library. The fix type depends on commercial value: high-priority prompts (P1) get a Create recommendation — How-To, Comparison, or Diagnostic depending on query intent. Lower-priority prompts may get Offsite Seed instead.
The 5 gap diagnosis dimensions.
When a strong match is confirmed, gap diagnosis compares your matched page to the competitor page currently being cited across five dimensions. The dimension that failed most determines the primary fix type assignment.
Freshness
How recently was your page updated relative to the typical decay window for this content type?
An outdated page loses citation share to fresher competitor content even when topic coverage is equivalent.
Structure and Extractability
Does your page have tables, numbered steps, definitions, or FAQ sections that AI engines can cite directly?
Pages without extractable structure are harder for AI engines to reference even when the underlying content is good.
Evidence Strength
Does your page carry the proof objects the competitor page has?
The strength scale runs from None → Weak → Moderate → Strong. Data, case studies, and third-party citations move the score up.
Audience Fit
Does your page target the same role, persona, and vertical as the query?
A page targeting a different buyer type fails this dimension even if the topic is correct.
CTA Quality
Does your page's CTA match what the prompt's query intent expects?
A comparison-intent query expects a trial or demo CTA; a how-to-intent query expects a guide or checklist. Mismatches fail this dimension.
A single failing prompt can produce a primary recommendation plus up to three secondary recommendations. The full fix type taxonomy is described on the Fix Type Taxonomy page .
How it works
Score
Every indexed page in your content library is scored against the unresolved prompt using all 7 signals.
Match
The match outcome (strong, borderline, or no match) routes the prompt to gap diagnosis or to a create recommendation.
Diagnose
Gap diagnosis compares your matched page to the currently-cited competitor page across 5 dimensions, producing a structured gap summary and a fix type assignment.
Content Library Matching powers the Diagnose engine — every Fix Type recommendation starts here.