
Answer Engine Optimisation: The Replacement For Keyword SEO
Search is moving from keyword-match to question-match. The page format that wins both.
Search is splitting. One half still returns a list of links and rewards whoever earns the click. The other half returns an answer and rewards whoever is cited inside it.
Optimising for the second is a different job, and most businesses are not doing it.
The unit changed
Traditional SEO optimises a page. Answer engine optimisation optimises a passage.
A model building an answer is looking for a self-contained chunk of text that resolves the question. It does not need your whole page. It needs forty to sixty words it can lift and attribute.
That single difference changes how the page should be written.
Put the answer first
Most business writing builds toward the point. Context, then framing, then the answer somewhere in the middle.
That structure fails here, because the extractable unit is buried inside material that does not stand alone. Put a direct answer at the top of the page, before any build-up, and write it so it makes sense with nothing around it.
Be specific enough to cite
Models favour text with things they can verify. Named standards, numbers, dates, technical terms, proper nouns.
Generic copy is structurally uncitable. There is nothing in it to attribute, so it gets passed over in favour of a source that said something checkable.
This is the largest gap in most commercial content. It is all adjectives.
Structure for extraction
Question-shaped headings, because the heading tells the model what the section answers. Short paragraphs, because a wall of text has no clean boundaries. FAQPage schema where the page genuinely answers questions.
None of that is a trick. It is writing that is easy to pull from, which is also writing that is easy to read.
Position works differently
In a list, being eleventh is worth almost nothing. In an answer there is no eleventh. There is cited or not cited, and typically a handful of sources get named.
That is the opening. You can be cited alongside brands you would never outrank, if you answered one specific sub-question better than they did.
The practical consequence: chase specific questions rather than broad terms. The broad term is contested. The specific question frequently is not.
Test it
Ask the AI tools the questions your buyers ask. See who gets cited and what shape their content takes.
Fifteen minutes of that teaches you more than any report, and almost nobody is doing it.
Written by David Eid. Published .
Read next.
Contact the Ignis Team
Send through your details and we will audit your business before we reply.




