RULE: Lead paragraphs with the direct answer. Aim for 40–60 words. Make every paragraph self-contained.

Rule: Every section of Candid client content leads with a direct, self-contained answer in 40-60 words, named entities and statistics where possible, before any setup or context. Each paragraph stands alone — a reader landing on it from a fragment URL or a chunked retriever understands it without prior context.

Why: The Princeton GEO paper (see Princeton GEO paper (Aggarwal et al., KDD '24) — the foundational generative engine optimization study) measured what AI engines actually extract:

• Quotation Addition: +41% on PAWC (GEO finding: Quotation Addition is the top-performing tactic at +41% on Position-Adjusted Word Count)
• Statistics Addition: +31% (GEO finding: Statistics Addition is the #2 tactic at +31%)
• Cite Sources: +28% (GEO finding: Cite Sources lifts visibility +28%)
• Keyword stuffing: -8 to -10% (GEO finding: Keyword stuffing is the only tested tactic that hurts AI visibility (-8% to -10%))

The 40-60 word target aligns with featured-snippet research showing 45-word paragraphs appear most frequently on SERPs. See Extractability: a quotable paragraph leads with the answer, is 40-60 words, lives under semantic HTML, and names entities concretely for the full synthesis.

How to apply:

• "The best X is Y" beats "Y might be a good option for X."
• Avoid "above" / "below" — chunkers strip surrounding context.
• Use proper nouns. Not "several companies" — name them.
• Carry a statistic or a verbatim quote in any paragraph where one exists.
• Each <h2> / <h3> is followed by a paragraph that fully answers the heading's implied question.