Say Hello

THE ANSWER-FIRST
PATTERN.

AI engines extract pages looking for direct answers near the top. The inverted-pyramid pattern (question, answer, context) is the format that wins extraction. Specific examples plus the anti-patterns that cost citations.

Pattern
Inverted Pyramid
First Answer
<60 words
Extraction Lift
Notable
Skill Required
None

HOW AI EXTRACTS.

When an AI engine processes a page for citation, it scans for content that directly answers the user's query. The scan is biased toward content near the top: above-the-fold paragraphs, opening sentences of <article> blocks, summary blocks. Content buried 2,000 words deep is rarely surfaced even when it is the better answer.

The pattern that wins extraction is the inverted pyramid: state the answer first, then provide context, then nuance. Journalism has used this for a century; AI extraction has revived it as a structural advantage.

THE PATTERN.

Three parts, in order:

EXAMPLE.

Anti-pattern (narrative open):

"In 2024, OpenAI began restructuring its crawler infrastructure. The change took place gradually over several months and reflected a broader trend in the AI industry toward separating training and search functions. Meanwhile, the SEO community was slow to react..."

Pattern (answer-first):

"OpenAI now operates three distinct crawler bots: GPTBot for training, OAI-SearchBot for the ChatGPT search index, and ChatGPT-User for on-demand fetches. Each has a different purpose and different correct robots.txt policy. Below is the breakdown."

Both versions can be the opening of the same article. The first is more readable as long-form prose. The second is dramatically more extractable - the AI engine reads the first three sentences and has the answer. Use the second.

WHERE IT DOES NOT APPLY.

The pattern is for informational and reference content. It is not for:

THE BOTTOM LINE.

Inverted-pyramid is the highest-leverage content pattern for AI extraction. Every research / how-to / reference article should default to it. Every long-form piece should have at least an answer-first summary block at the top, even if the body is narrative. Cost to implement: zero. Payoff: durable extraction lift across every AI surface.

Stop Guessing What AI Sees

MEASURE THE LEVERS
THAT ACTUALLY EXIST.

If you want this methodology applied to your specific site - your real logs, your real citation data, your real fix list - the audit is the productized way to do it.