Content analyzer
What is the citation funnel (Retrieval, Citation, Overall)?
The citation funnel is Seeklyn's 3-stage content score, each rated 0–100: Retrieval measures whether a retrieval index can find your passages for the target query, Citation measures whether the retrieved passages are actually quotable, and Overall is the mean of the two. A passage has to clear both stages to be cited — found first, then quoted.
The funnel appears at the top of every content report in Seeklyn — run a page free at /dashboard/content.
What is the citation funnel?
The citation funnel is Seeklyn's 3-stage content score that models the 2 things that must both happen before an AI engine cites a passage: retrieval (can the index find it?) and citability (is it worth quoting?). Each stage is scored 0–100. Overall is the mean of Retrieval and Citation, so a page with Retrieval 80 and Citation 40 scores Overall 60 — not 80.
The funnel is ordered because retrieval comes first: a passage never retrieved can never be cited, regardless of how well written it is. Seeklyn evaluates the 2 stages in sequence and reports all 3 numbers side by side so you can see which half is the bottleneck.
What does each stage measure?
Each stage measures a distinct question, and the 3 numbers are always shown together. Retrieval asks whether the passage can be found, Citation asks whether it can be quoted once found, and Overall averages the two into a single headline number.
| Stage | Range | Question it answers | How it is measured |
|---|---|---|---|
| Retrieval | 0–100 | Can a retrieval index find your passages for the query? | BM25 over your own chunks; passages with zero query-term overlap are orphaned |
| Citation | 0–100 | Are the retrieved passages actually quotable? | Facts, attribution, answer-first structure, and self-containment per passage |
| Overall | 0–100 | How does the page do end to end? | The mean of Retrieval and Citation |
How is the Overall score calculated?
Overall = (Retrieval + Citation) ÷ 2 — exactly 2 numbers averaged into 1. Because it is a simple mean, Overall stays mode-invariant: the same content and query always produce the same Overall regardless of which analysis mode you pick. The number is comparable across runs, so you can track progress with confidence.
Retrieval is monotonic: a passage with zero query-term overlap (BM25 = 0) is orphaned and scores 0, and adding query-relevant terms can only raise that score. Citation is monotonic under edits as well: applying a fix that clears an extractability flag must never lower a comparable citability score — Seeklyn enforces this invariant on every rewrite.
Frequently asked questions
Why separate Retrieval from Citation?
Because they fail for different reasons and take different fixes. A passage can be perfectly quotable yet never retrieved (zero query overlap), or easily retrieved yet too vague to quote. Splitting the funnel into 2 stages tells you which problem you actually have.
What does an Overall of 70 mean?
An Overall of 70 is the mean of your Retrieval and Citation scores — for example Retrieval 82 and Citation 58 average to 70. Read the 2 underlying stages to see whether the gap is being found or being quotable, then fix the lower one.
Can the Overall score change without editing the page?
The Overall stays stable for the same passages and query — it is mode-invariant, so switching analysis modes will not move it. It changes when you edit the content, change the target query, or apply a fix that clears an extractability flag.