Why is the passage the unit of competition now and how to write for it?

What does it mean that the passage is the unit of competition?

The passage is now the unit of competition because AI crawlers extract and rank snippets, not whole pages. CRO9’s data shows ~88% of AI Overview appearances are informational‑intent queries, meaning users expect a direct answer. When a crawler scans a page, it isolates 200‑400 word blocks that fully answer a query and treats each block as a separate ranking entity. This shift explains why classic page‑rank signals predict only ~38% of AI citations. To win, you must write passages that are self‑contained, answer‑focused, and positioned within the top 30% of the page—where 55% of citations originate. By structuring content into clear, answer‑first sections, you increase the chance that an AI will surface your passage in the answer box, driving higher CTR and brand visibility.

How long should each passage be for optimal AI citation?

Optimal AI citation length is 200‑400 words per passage. CRO9’s tracker defines this range as the “unit of citation,” and data shows passages within this window capture the majority of citations while staying under the keyword density threshold of 1.5‑2%. Longer blocks dilute relevance and risk keyword stuffing, which can penalize rankings. Shorter blocks may not provide enough context for the AI to deem the answer authoritative. By keeping each section concise and focused, you align with the AI’s parsing logic and improve the likelihood of appearing in the answer box, especially since 40‑60% of AI citations churn monthly and only the strongest passages persist.

Which on‑page signals boost a passage’s chance to be cited?

On‑page signals that lift passage citations include placement near the top of the page, clear heading hierarchy, and fast Core Web Vitals. CRO9’s data indicates 55% of citations pull from the top 30% of the page, so positioning key passages early is critical. Additionally, meeting CWV targets—LCP under 1.8 s, INP under 150 ms, CLS under 0.05, and TTFB under 400 ms—ensures the crawler renders the passage quickly, avoiding the “BLOCKED” or “RENDERS‑EMPTY” states CRO9 detects across six AI bots. Finally, maintaining keyword density below 2% and using natural language improves AI relevance scoring, as overly optimized passages are filtered out.

How can I measure the impact of passage‑focused writing?

Measure impact with CRO9’s 8.2 KB gzipped tracker, which records 28 distinct visitor‑behaviour events, including scroll depth and citation clicks. Track the number of AI citations per passage and compare against baseline page‑level metrics. CRO9’s AI‑crawler monitoring distinguishes between blocked requests and empty renders, letting you see if a passage is being indexed. Look for a rise in branded query CTR (+18% on average) and a reduction in bounce rate when passages rank in answer boxes. Combine this with Core Web Vitals reports to ensure performance isn’t throttling citation potential.

What writing techniques ensure a passage earns AI citations?

Effective techniques include starting each passage with a direct answer, using a single focus keyword at <2% density, and embedding verifiable data points. CRO9 recommends structuring each 200‑400 word block as a mini‑article: answer first, then provide supporting facts, and close with a concise takeaway. Use bullet‑point key facts (as shown above) to highlight numbers—AI models favor quantifiable data. Keep sentences under 20 words to aid parsing, and avoid filler introductions. By delivering a self‑contained, data‑rich answer, you align with the AI’s preference for concise, authoritative snippets that earn citations.