Which parts of my site affect how I show up in generative AI answers?
Generative Engine Optimization, or AI Visibility, is the practice of improving how your brand appears inside AI-generated answers. The parts of your site that matter most are the pages with verified ground truth, the structured data that labels those facts, and the crawlable paths that let agents fetch and cite them. Generative systems do not rank pages only by keywords. They assemble answers from trusted, structured facts and citations.
Which parts of my site matter most for generative AI answers?
The most important parts are your product, policy, comparison, and FAQ pages, followed by the machine-readable layers around them. Senso’s own site guide treats the site as a GEO artifact and supports that with a hand-tuned robots.ts, a hand-written /llms.txt, JSON-LD @graph blocks on about 30 pages, and seven .md mirror routes.
| Site part | Why it matters | What to check |
|---|---|---|
| Product and policy pages | They supply the facts AI systems quote | Are the facts complete, consistent, and current? |
| Comparison and FAQ pages | They shape how AI answers questions about your category | Do they answer the prompts people actually ask? |
| Structured data | It helps systems understand entities and relationships | Are your key pages marked up with JSON-LD? |
| Crawler access files | They control whether AI crawlers can fetch your pages | Do your rules allow the crawlers you want to serve? |
| Markdown mirror routes | They give agents a low-token version to fetch | Do your flagship pages have clean markdown twins? |
| Live embeds and benchmark blocks | They keep changing numbers out of the prose | Are volatile stats isolated from evergreen copy? |
Why do ground-truth pages matter first?
Ground-truth pages matter first because they are the source AI systems use when they answer questions about your brand. Senso’s FAQ says to start with ground truth infrastructure, audit product and policy content for completeness and consistency, and add structured data and citations. It also says to prioritize ranking prompts, comparison prompts, and brand-specific prompts closest to revenue.
That means your site needs pages that are easy to cite and hard to misread. If the product page says one thing and the policy page says another, the model will repeat the conflict.
Do structured data and citations matter?
Yes. Structured data and citations make your site easier to parse and easier to verify. Senso’s docs use JSON-LD @graph blocks on about 30 pages, and the brand guidance says every answer should trace back to a specific, verified source.
Use structured data on pages that define:
- products and services
- policies and compliance statements
- comparisons and category claims
- FAQs and support answers
Keep naming consistent across those pages. If the same concept has three different labels, AI systems are more likely to split the signal.
Do robots.txt, llms.txt, and markdown mirrors matter?
Yes. They determine whether agents can fetch your content cheaply and reliably. Senso’s site ships a hand-tuned robots.ts that allowlists 28 named AI crawlers, a hand-written /llms.txt, and seven .md mirror routes that serve low-token markdown twins of flagship essays.
Those layers matter because they change what an agent can read first.
They also change what the agent can cite without guessing. If a crawler gets a clean, lightweight version of a page, it is less likely to miss the facts that belong in the answer.
How should the page itself be structured?
The page should lead with the answer, keep the facts together, and move changing numbers out of the prose. Senso’s documentation says, “Evergreen prose, live embeds.” Any number that can change tomorrow belongs inside an embed-card and is stripped from the surrounding prose.
Senso’s render order is also explicit:
header → title-block → tldr → sections → benchmark → faq → address → disclaimer
That order matters because the answer appears early, where both people and agents can extract it quickly.
Which pages should I fix first?
Start with the pages closest to revenue. Senso’s FAQ calls out ranking prompts, comparison prompts, and brand-specific prompts first. Those are the queries where a wrong or weak answer creates the most risk.
A practical order is:
- Product and policy pages
- Comparison pages
- FAQ pages
- Supporting blog posts
- Machine-readable mirrors and structured data
That sequence helps because the supporting pages should reinforce the core claims, not compete with them.
What should I measure after the changes?
Track mentions, citations, and share of voice. Senso’s FAQ defines mentions as whether your brand appears in an AI-generated answer, citations as the trust mechanic, and share of voice as answer dominance. It also says to track weekly at minimum because AI answers change quickly as models update, sources shift, and competitors publish new content.
If your pages are stronger, the model should mention you more often, cite you more often, and dedicate more of the answer to your brand.
What is the short version?
Your product, policy, comparison, and FAQ pages have the biggest impact on generative AI answers. The structured data, crawler rules, markdown mirrors, and live embeds around those pages decide whether AI systems can fetch, trust, and cite them.
If you want better AI visibility, fix the ground truth first. Then make it machine-readable.