Content analyzer
What is a run-on sentence (RUN_ON) in AEO content?
A sentence exceeding 30 words (strings.Fields count) fires Seeklyn's RUN_ON flag, because packing multiple ideas into 1 sentence prevents AI engines from extracting a single atomic fact. 15–25-word sentences are the clearest for AI extraction; RUN_ON sentences score lower on citability but remain "liftable" unless a second flag also fires.
Seeklyn flags this in every content analysis — run your page free at /dashboard/content.
What is a run-on sentence in AEO?
A run-on sentence is any sentence that packs multiple ideas, clauses, or facts into a single syntactic unit — making it impossible for an AI extractor to quote it as a single atomic fact. Seeklyn's RUN_ON flag fires when a sentence's word count (measured by strings.Fields) exceeds 30 words. AI engines retrieve passages sentence-by-sentence; a 40-word sentence that contains 3 separate claims cannot be cleanly cited for any one of them.
Run-on sentences reduce citability without breaking self-containment. Unlike ANAPHORA_OPEN and NO_ANSWER_FIRST — which set self-contained = false — RUN_ON leaves the sentence verdict as "liftable" unless another flag also fires on the same sentence. The sentence is extractable in principle, but its density makes it a lower-priority candidate for an AI engine choosing between multiple citable passages.
How does Seeklyn detect run-on sentences?
RUN_ON fires in Seeklyn's sentenceFlags() function when len(strings.Fields(s)) > 30. RUN_ON is a per-sentence word-count check: strings.Fields splits on whitespace, so hyphenated compounds count as 1 word each. The threshold is 30 — a sentence with exactly 30 words does not fire; a sentence with 31 does.
RUN_ON is classified as a moderate-severity flag. RUN_ON does not flip the self-contained badge and does not affect whether a sentence is anaphoric or answer-first. Its weight in the 0–100 citability score is lower than the 2 structural flags but comparable to HEDGE. A passage with many run-on sentences can still be self-contained and extract cleanly — it simply scores lower on the citability scale than an equivalent passage rewritten into concise sentences.
How to fix a run-on sentence for AI citation
Split at a natural clause boundary and give each sentence 1 claim. Before: "While structured data has long been used to help search engines understand page content, it is now also used by AI answer engines like Google AI Overviews and ChatGPT to identify citable passages, which means that marking up your content with schema.org types can now directly influence whether your page gets cited." After: "Structured data now directly influences AI citation. Schema.org markup helps Google AI Overviews and ChatGPT identify citable passages — a role it shares with classic search indexing."
15–25-word sentences are the clearest for AI extraction. Each sentence should carry 1 claim, 1 subject, and ideally 1 concrete fact. When splitting a run-on, check that each resulting sentence also clears the other flags: the first sentence should pass answersFirst(), neither sentence should open with an anaphoraOpener, and any qualifying claims should be stated declaratively rather than with a hedge word.
Frequently asked questions
Is 30 words a hard limit?
Yes in Seeklyn's rubric — a sentence with 31 words fires RUN_ON. The threshold is a practical citability heuristic backed by readability research showing that sentences over 25–30 words reduce comprehension speed. 30 is the exact boundary in Seeklyn's sentenceFlags() implementation.
Does a run-on sentence orphan the whole passage?
No — RUN_ON reduces the sentence's extractability score but the verdict stays "liftable" unless ANAPHORA_OPEN or HEDGE also fire on the same sentence. Only ANAPHORA_OPEN triggers the "orphaned" verdict. A passage of run-on sentences is self-contained but low-scoring on citability.
What Flesch-Kincaid reading level does a 30-word sentence correspond to?
A single 30-word sentence pushes the Flesch-Kincaid grade level well above 12 for that clause. Seeklyn surfaces Flesch-Kincaid readability separately as its own score — both scores often improve together when you split long sentences into 15–25-word units.
Can a passage with many run-on sentences still be self-contained?
Yes — self-containment is determined only by ANAPHORA_OPEN and NO_ANSWER_FIRST. A passage full of run-on sentences is self-contained (self-contained = true) but low-scoring on the 0–100 citability scale. Clearing RUN_ON flags is the right next step once the structural flags are resolved.