How to Get Your Brand Named in AI Answers (Not Google Rankings)
Short answer: To show up in AI answers, build pages that answer buyer questions with citable facts — then make sure machines can retrieve them (crawl, language, structure). Blue-link Google rankings help, but they aren’t the only gate: many AI answers pull from pages that are clear, consistent, and third-party reinforced — not from whoever sat #1 yesterday.
What this is not
- Chasing ten “GEO magic” backlinks with no answer page
- Pasting English FAQ into an Indonesian market
- Assuming search ads equal citations in AI Overview / Gemini
llms.txt won't put your brand in AI answers
An llms.txt file alone won’t put a brand in AI answers; Google says AI Overviews and AI Mode need no AI text file or special markup. White Wood’s own monitoring engine, SOLEDAD, found whitewood.ai named in 2 and cited in 4 of 25 Google AI Overviews answers for “cara agar brand muncul di jawaban AI” from 18 September to 18 October 2026, and absent on the other five engines it tracks.
llms.txt is a proposed Markdown file that points AI agents to the most useful pages on a site. Jeremy Howard published the proposal in September 2024, and its site says it is used most heavily for software documentation, where coding agents follow it to find API references (llmstxt.org). That is a real use. It is not the same as asking a search engine to cite you in an answer.
Google’s guidance for site owners says there are no additional requirements to appear in AI Overviews or AI Mode (Google Search Central). To show as a supporting link, a page must be indexed and eligible for a snippet, the same as normal Search. If a vendor’s plan for Google’s AI answers starts with llms.txt, ask what happens to your pages.
Add llms.txt when agents need to use your site, not when you want to be recommended. Documentation, product specs, policies and price lists are good candidates. It is a small job and does no harm. Whether a given assistant reads your file when deciding whom to recommend is not documented, so don’t report it to the board as AI visibility progress. Track the same prompts before and after any change instead, as in the 90-day order below.
What this cannot tell you. Google’s guidance covers Google’s own AI features. It says nothing about how other assistants treat llms.txt, and getting the basics right is no promise of a citation. Those counts cover one month on six engines and do not show which passage of a page an answer used.
Declared interest: White Wood sells this work and measures it with SOLEDAD, our own AI-monitoring engine.
Sources: Google Search Central, AI features and your website; llmstxt.org (Jeremy Howard, 3 September 2024); SOLEDAD baseline for whitewood.ai on six engines, 18 September to 18 October 2026.
What usually works
1. Answer unbranded questions first.
“How to measure…”, “services in Jakarta”, “X vs Y”. Pages that only praise themselves rarely get cited for category asks.
2. One stable fact pin.
Address, hours, price bands, service definitions, process — written the same across the site. Contradictions between pages = the model picks a cleaner competitor.
3. Structure humans and machines can both read.
Answer up top, short facts, then Q&A that repeats natural questions. Narrative Architecture (White Wood): arrange the page so it can be cited — not jargon stuck in a footer.
4. Local corroboration.
In Indonesia, citations from trusted media or lists often beat a quiet brand blog. One owned page + one aligned outside source > ten mill articles.
5. Technical retrievability.
Don’t block Search crawlers while complaining “AI never names us.” robots, CDN rules, and a blunt “block AI” default can close the door you need.
A practical 90-day order
- Freeze 20–40 buyer prompts
- Measure baseline (appears / not + sources)
- Pick five pages that close the biggest gaps
- Rewrite answer-first (market language)
- Measure again — change pages, not only the dashboard
Straight answers
Is SEO dead?
No. What’s changed: click traffic isn’t the only score. Citation inside answers is a new surface.
Is homepage optimization enough?
Rarely. Specific questions need specific pages.
Is this the same as “ChatGPT ads”?
No. Organic mention in answers ≠ buying a slot inside chat.
