Quick Answer: AAO (AI Agent Optimisation) is the practice of making your business discoverable and actionable to autonomous AI agents that browse, research, and transact on behalf of users. In Singapore, this means publishing a /llms.txt file, implementing machine-readable JSON-LD schema, and structuring your digital presence so AI agents can read, evaluate, and act on your business information without friction.
Something is researching your competitors right now
Somewhere in Singapore, a potential customer has just asked their AI assistant to find the best logistics company for their next shipment. The AI does not return ten blue links. It browses. It reads company pages. It compares service details. It shortlists three options and drafts a message to each.
If your business is readable to that agent, you are on the shortlist. If it is not, you are invisible — and you will never know the opportunity passed. This is the world AAO was built for.
What is AAO?
AAO stands for AI Agent Optimisation (also called Assistive Agent Optimisation). It is the fourth and newest discipline in digital search optimisation — after SEO, AEO, and GEO.
Where the first three disciplines optimise for how humans discover businesses (through search results, AI answer boxes, and LLM recommendations), AAO optimises for how autonomous AI agents discover, evaluate, and interact with businesses on a user's behalf.
An AI agent is not a chatbot that waits for questions. It is an autonomous system given a task — "find me a reliable freight forwarder in Singapore" — that it completes by browsing, reading, comparing, and acting. The agent makes decisions. The human reviews the results.
For AAO to work in your favour, your business must be:
- Discoverable — AI agents can find your site and identify what you do
- Readable — Your services, pricing signals, and contact details are structured for machine access
- Actionable — The agent can extract enough information to take a next step (send an enquiry, request a quote, add you to a shortlist)
Why AAO matters in Singapore in 2026
Singapore has one of the highest AI adoption rates in Asia Pacific. GovTech is rolling out AI assistants to 150,000 public sector officers. Enterprise teams across finance, logistics, healthcare, and professional services are deploying autonomous agents for procurement research, vendor evaluation, and service comparison.
The consumer side is moving just as fast. AI assistants embedded in smartphones, productivity tools, and browsers are gaining the ability to complete tasks — not just answer questions.
The businesses that are readable to these agents today are building a discovery advantage that will compound as agent adoption accelerates. The businesses that are not readable will lose an increasing share of top-of-funnel discovery to competitors who are — without any change in their Google rankings to signal the shift.
The three layers of AAO
Layer 1: llms.txt — your machine-readable introduction
/llms.txt is a plain-text file published at the root of your website that tells AI agents and language models who you are, what you do, what pages exist on your site, and how to contact you.
Think of it as a robots.txt — but instead of directing search engine crawlers, it directs AI agents and LLMs toward the information they need to understand and represent your business accurately.
A well-structured llms.txt file should include: company name, description, and slogan; core services with one-line descriptions; key page URLs with brief descriptions; contact details (email, phone, location); and notes for AI agents (pricing availability, how to enquire, what the business does not do).
InfinitusNow publishes its own llms.txt at infinitusnow.com/llms.txt — maintained as a working example of the AAO practices we implement for clients.
Layer 2: JSON-LD schema — structured data for machine understanding
Schema markup is the language Google, AI engines, and autonomous agents use to understand what a webpage means — not just what it says. For AAO purposes, the most important schema types are:
Organization — Tells agents your business name, description, founding date, contact details, social profiles, and service area. This is the foundation that makes your business a recognised entity in AI systems.
ProfessionalService / LocalBusiness — Specifies your service category, geographic coverage, and operational details. Agents researching "logistics companies in Singapore" use this to filter and compare.
Service — Describes individual services with names, descriptions, and pricing signals. An agent comparing freight forwarders needs to understand what each company actually offers.
FAQPage — Structures your FAQ content for direct extraction by AI systems. Agents researching your category will pull from FAQ schema to answer their user's questions.
ContactPoint — Makes your phone, email, and enquiry channels explicitly machine-readable. An agent completing a task on a user's behalf needs to know exactly how to reach you. See our JSON-LD schema guide for full implementation details.
Layer 3: WebMCP / API endpoints — agent-transactable access
The frontier of AAO is structured API endpoints that allow AI agents to take direct actions — retrieving your service catalogue, checking availability, or submitting an enquiry — programmatically, without scraping a webpage.
The emerging WebMCP standard (Model Context Protocol) defines how AI agents should discover and interact with web-accessible tools. Businesses that implement MCP endpoints give agents a structured, authoritative channel to interact with them — a significant advantage as agentic capabilities expand.
InfinitusNow's own site exposes MCP endpoints at /.well-known/mcp.json, making it one of the first Singapore digital agencies to implement WebMCP. This is the standard we implement for clients in the Dominator tier. See our WebMCP explainer for the full breakdown.
AAO vs SEO vs AEO vs GEO
| SEO | AEO | GEO | AAO | |
|---|---|---|---|---|
| Optimises for | Search rankings | AI Overview answers | LLM brand citations | AI agent discovery |
| Discovery surface | Google/Bing results | Featured snippets, AI Overviews | ChatGPT, Perplexity, Gemini responses | AI assistants, agent browsers |
| User action | Clicks a link | Reads the answer | Reads the recommendation | Delegates research to AI |
| Key signals | Keywords, backlinks, technical health | FAQ schema, concise answers, structured headings | Entity authority, off-site citations, brand consistency | llms.txt, JSON-LD schema, MCP endpoints |
| Result | Page ranking | Citation in AI answer | Brand recommendation | Agent shortlisting |
| Timeline to results | 3–12 months | 4–8 weeks | 4–12 weeks | Immediate (on deployment) |
How to implement AAO for your Singapore business
Step 1: Publish a /llms.txt file (today)
This is the single highest-impact, lowest-effort AAO action. Write a structured plain-text file and publish it at yourdomain.com/llms.txt. Include your business description, services, key pages, and contact details.
Step 2: Audit your JSON-LD schema (week 1)
Use Google's Rich Results Test to check what structured data your site currently publishes. At minimum, every Singapore business should have Organization and LocalBusiness schema with complete, accurate information.
Step 3: Add Service schema to service pages (week 1–2)
For each core service, add a Service schema block that describes the service name, description, and provider. This is what agents use to compare your offering against competitors.
Step 4: Implement FAQPage schema on key pages (week 2)
Convert your existing FAQ sections into properly structured FAQPage JSON-LD. This makes your Q&A content directly extractable by AI agents.
Step 5: Check robots.txt for AI crawlers (week 1)
Ensure your robots.txt does not inadvertently block AI crawlers. Some AI crawlers use different user-agent strings than traditional search bots — confirm that Anthropic's ClaudeBot, OpenAI's GPTBot, Google's Gemini crawlers, and Perplexity's crawler are all permitted.
Step 6: Reference /llms.txt in robots.txt (week 1)
Add a Llms-txt: directive to your robots.txt pointing to your llms.txt file. This signals to AI systems that the file exists and should be consulted.
The AAO opportunity in Singapore
Of the four DIP disciplines, AAO currently has the thinnest competition among Singapore businesses. Most agencies have published blog posts about AEO and GEO. Very few are implementing AAO technically — and almost none are offering it as a managed service.
This creates a genuine first-mover window. Businesses that implement AAO foundations now — before agent-mediated discovery becomes mainstream — build a structural advantage that is very difficult for late movers to close.
InfinitusNow is one of the few Singapore agencies implementing AAO on its own properties and for clients. Our /llms.txt, MCP endpoints, and schema implementation exist as working proof — not just a service page.
AAO in the DIP framework
In InfinitusNow's Digital Intelligence Platform, your AAO Score is one of four components in your overall DIP Score. It measures: /llms.txt existence, completeness, and accuracy (30%); JSON-LD schema coverage and quality (30%); robots.txt AI crawler configuration (15%); WebMCP or API endpoint availability (15%); sitemap currency and structure (10%).
A perfect AAO Score requires all five components to be implemented correctly and kept current. Most Singapore businesses starting from Day Zero score 0–15 on AAO — because the discipline barely existed as a formalised practice before 2025.
Frequently asked questions
What does AAO stand for?
AAO stands for AI Agent Optimisation (also referred to as Assistive Agent Optimisation). It is the practice of making a business discoverable, readable, and actionable to autonomous AI agents that research and transact on behalf of users.
Is AAO different from SEO?
Yes, substantially. SEO optimises for human users who choose to click search results. AAO optimises for AI agents that autonomously browse, compare, and act on a user's behalf — often without the user seeing the intermediate steps. The two disciplines share some foundations (structured content, technical site health) but diverge significantly in implementation.
What is llms.txt?
llms.txt is a plain-text file published at the root of your website that tells AI agents and language models who you are, what you offer, what pages exist on your site, and how to contact you. It functions similarly to robots.txt — but for AI systems rather than search engine crawlers.
Do AI agents actually visit websites in Singapore?
Yes. Multiple AI platforms deploy autonomous agents that browse the web to answer user queries, complete research tasks, or gather information for agentic workflows. This includes agents from Anthropic, OpenAI, Google, Perplexity, and others. Singapore's high AI adoption rate means this activity is accelerating faster here than in most markets.
How quickly do AAO improvements take effect?
Unlike SEO (which can take months to reflect in rankings), AAO improvements take effect as soon as AI crawlers re-index your site — typically within days to weeks of publication. Publishing a /llms.txt file and implementing schema markup are among the fastest-acting improvements in the DIP framework.
Is AAO only relevant for large businesses?
No. AAO is especially valuable for SMEs, because it levels the playing field. An AI agent evaluating vendors does not favour the business with the largest budget or the most backlinks — it favours the business whose digital presence is the most structured, accurate, and readable. A well-implemented AAO foundation gives a small agency the same agent-discoverability as a large one.