How to make your website AI-ready in Malaysia (2026)

An AI-ready website is one an assistant can fetch, read and quote without running any code. That matters because none of the major AI crawlers execute JavaScript, so a page that assembles itself in the browser can arrive at ChatGPT, Claude or Perplexity as an empty shell (Vercel and MERJ, 2024). Most Malaysian business sites need targeted fixes, not a rebuild.

Key takeaways

• Vercel and MERJ analysed AI crawler traffic across their network and concluded that "none of the major AI crawlers currently render JavaScript", naming OpenAI's OAI-SearchBot, ChatGPT-User and GPTBot, Anthropic's ClaudeBot, Meta-ExternalAgent, ByteDance's Bytespider and PerplexityBot (Vercel, December 2024).
• Gemini is the exception. It "leverages Googlebot's infrastructure, enabling full JavaScript rendering", which is why one business can be visible in Gemini and invisible in ChatGPT at the same time (Vercel, December 2024).
• Schema markup is oversold as a citation lever. Ahrefs tracked 1,885 pages that added JSON-LD against 4,000 control pages and measured Google AI Overviews at −4.6%, AI Mode at +2.4% and ChatGPT at +2.2% (Ahrefs, May 2026).
• Google states plainly that "[s]tructured data isn't required for generative AI search, and there's no special schema.org markup you need to add" (Google Search Central).
• The audience is already there. Ipsos found 72% of Malaysians use AI tools despite not always trusting them (Ipsos AI Monitor 2026).

New to this? Acclaira runs a free one-hour AEO class on Zoom, every 2 weeks on a Thursday at 9.30pm Malaysia time. Ask for the Zoom link on WhatsApp.

What does an AI-ready website actually mean?

An AI-ready website is one whose meaning survives being read as plain text by a machine that will not run scripts, will not click, and will not wait. Three things decide it: the crawler is allowed in, the words are present in the raw HTML that comes back, and the page states plainly who you are, where you operate and what you sell.

Being online is not the same as being readable. The Department of Statistics Malaysia reported that 95.3% of establishments had internet access in 2024, up from 94.0% the year before, and that web presence reached 74.4%, up from 72.7% (DOSM, 2026). Roughly three in four Malaysian establishments already have a web presence, so the competitive question is no longer whether you have a site. It is whether an assistant can read the one you have.

Google's own guidance points the same way. It tells site owners to "ensure your content is crawlable, as Google Search generative AI models use publicly accessible, crawlable content" (Google Search Central). Our comparison of AEO and SEO for Malaysian businesses sets out where the two disciplines actually diverge.

Can ChatGPT actually read my website right now?

Often it cannot, and the reason is JavaScript. Vercel and MERJ studied AI crawler behaviour across Vercel's network and found that the major AI crawlers fetch JavaScript files but never execute them, so any content painted into the page by the browser is invisible to them. GPTBot requested JavaScript files in 11.50% of its requests and ClaudeBot in 23.84%, without running a single one (Vercel, December 2024).

This is the single most expensive misunderstanding we meet in Malaysian SME websites. A site built as a client-rendered app, or a page whose text lives inside a script-driven slider, tab or accordion, can look perfect to you and return almost nothing to the crawler. The volumes involved are not small: in the month Vercel measured, GPTBot made 569 million fetches and Claude 370 million, against Googlebot's 4.5 billion (Vercel, December 2024).

Source for all rows: Vercel and MERJ, "The rise of the AI crawler", 17 December 2024.

Permission is the other half. OpenAI runs three crawlers with three different jobs, and only one of them governs whether you can appear in ChatGPT's search answers. OpenAI states that OAI-SearchBot is "used to surface websites in search results in ChatGPT's search features", that "[s]ites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers", and that ChatGPT-User "is not used to determine whether content may appear in Search" (OpenAI). Blocking the training crawler and blocking the search crawler are separate decisions, and many website platforms collapse them into one checkbox.

How do I check my own site in five minutes?

Ask your browser for the page without the browser's help. In Chrome, open your site, press Ctrl+U to view the page source, then press Ctrl+F and search for a distinctive sentence from your homepage. If the sentence is in that raw source, ChatGPT can read it. If it is missing, you are looking at your JavaScript gap. Then open yourdomain.com/robots.txt and check whether OAI-SearchBot appears under a Disallow rule.

Two checks, five minutes, no developer. Neither one tells you whether an assistant will recommend you, which is a separate test covered in our guide to checking whether ChatGPT recommends your business.

Do I need to rebuild my website, or can my current one be fixed?

Most sites can be fixed. A rebuild is only the honest answer when the platform itself prevents the fixes, when you cannot edit the raw HTML or the robots.txt, or when the cost of working around the platform exceeds the cost of leaving it. Everything else on the AI-readiness list is content and configuration work you can do on the site you already own.

Google is unusually direct about this, which is useful when a vendor tells you otherwise. Its guidance states: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search" (Google Search Central). Treat any quotation that begins with a mandatory rebuild as a claim that needs evidence.

Budget sits under this decision, so read it with real numbers beside it. Our guide to what a Malaysian SME website should cost sets out the ranges, and our comparison of Instagram, WhatsApp and a real website covers the case where the honest answer is that your site is not where your customers meet you at all.

What is schema markup, and does my business really need it?

Schema markup is structured data you add to a page, usually as JSON-LD, that states in a fixed vocabulary what the page is about: the business name, address, opening hours, the author of an article, the questions in an FAQ. It helps machines classify a page. What it does not reliably do is buy you AI citations, and the evidence on that is now specific.

Ahrefs identified 1,885 pages that added JSON-LD schema between August 2025 and March 2026, matched them against 4,000 control pages, and measured citation change before and after. Google AI Overviews moved −4.6%, AI Mode +2.4% and ChatGPT +2.2% (Ahrefs, May 2026). The authors' own summary is the fairest reading of it:

"For pages already getting cited by AI, adding JSON-LD schema didn't boost citations on Google AI Mode or ChatGPT, and showed only a small decline in AI Overviews that we can't clearly attribute to schema."

— Louise Linehan and Xibeijia Guan, Ahrefs, We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved. (11 May 2026)

So add schema, but for the right reasons. Ahrefs still holds that it earns its place for "rich results, voice assistants, knowledge graphs, downstream entity recognition" (Ahrefs, May 2026), and Google confirms structured data "isn't required for generative AI search" (Google Search Central). At Acclaira we mark up LocalBusiness, Organization and FAQPage as basic hygiene, and we do not sell it as the thing that gets you named. The same guidance closes off a popular shortcut: Google says files such as llms.txt "will neither harm nor help your site's visibility or rankings in Google Search, as Google Search ignores them" (Google Search Central).

How should my pages be written so an assistant can lift the answer?

Write each page so any single paragraph can be quoted on its own. An assistant does not read your page top to bottom and summarise it. It retrieves passages. A paragraph that opens with the answer, names the business or the place explicitly, and carries a figure with a source attached is a paragraph that can be lifted without editing.

This is measured, not folklore. The Princeton-led GEO study built GEO-bench, a benchmark of diverse user queries across multiple domains, and reported that its creator-side optimisation methods "can boost visibility by up to 40% in generative engine responses" (Aggarwal et al., KDD 2024). The same paper found the effect varies by domain, so read 40% as a demonstrated ceiling on a benchmark rather than a number to expect on your own site. Question-shaped headings help for the same reason, because they match the shape of what people type into an assistant.

The practical rule for a Malaysian business is to name things an engine can verify. Say "Klang, Selangor" rather than "the Klang Valley area". Name the service in the words a customer would use. Give the price range if you can defend it, and the reason if you cannot. Our guide to why a business does not show up in ChatGPT or Google AI works through the failure modes behind this.

What decides the answer that is not on your website at all?

A large part of it sits off your domain, in the places an engine uses to corroborate that you exist. Google says local results "are mainly based on relevance, distance, and popularity", and that prominence draws on "how many websites link to your business and how many reviews you have" (Google Business Profile Help). Your website cannot supply any of that on its own.

Reviews are the clearest lever here, and Google is explicit that "[m]ore reviews and positive ratings can help your business's local ranking" (Google Business Profile Help). Ask real customers, never gate or incentivise the request, and keep your name, address and phone number identical everywhere they appear. Our article on whether Google Business Profile matters for AI search covers that surface in full.

In what order should a Malaysian SME do this?

Do the permission and rendering work first, because nothing downstream counts until an assistant can read the page. Content comes second, structured data third, and off-page corroboration runs alongside all of it. In our client work at Acclaira the first two steps are where the largest share of the recoverable problem sits, which is a practitioner observation rather than a published finding.

1. Check robots.txt and unblock OAI-SearchBot if it is disallowed.
2. Confirm your key pages' text appears in view-source, and fix the rendering if it does not.
3. Write the business facts plainly on the page: services, service area, hours, contact.
4. Rewrite the top pages answer-first, with question-shaped headings.
5. Add LocalBusiness, Organization and FAQPage schema as hygiene.
6. Build reviews and consistent listings off-site, patiently.

How do I know whether any of it worked?

Measure by asking the assistants the questions your customers ask, on more than one engine and more than one day, and recording whether you are named, mentioned or absent. Website analytics will not show this, because the assistant answers without sending a click. The baseline is the record you keep before the work starts.

Our own probe shows why the gap is worth attacking in Malaysia. On 1 August 2026, Acclaira asked ChatGPT, Gemini and Claude both of the questions this article answers, with web search enabled and the location set to Malaysia where the engine's API allows it. The three engines returned 100 cited sources across 72 unique domains. Exactly one was a Malaysian domain, an agency page, and ChatGPT returned no sources at all on either question. The answers were generic global checklists. That is an open field, not a crowded one.

For the method itself, our step-by-step guide to checking whether ChatGPT recommends your business sets out the prompts and the scoring, and ranking number 1 on Google but not appearing in ChatGPT explains why your Google position is not the measure to watch.

Prefer to be walked through it? Acclaira's free one-hour AEO class covers exactly this, every 2 weeks on a Thursday at 9.30pm Malaysia time on Zoom. Message us on WhatsApp to join.

Sources

• Giacomo Zecchini, Alice Alexandra Moore, Malte Ubl and Ryan Siddle, "The rise of the AI crawler", Vercel (with MERJ), 17 December 2024
• Louise Linehan and Xibeijia Guan, "We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved.", Ahrefs, 11 May 2026
• Google Search Central, "Google's Guide to Optimizing for Generative AI Features on Google Search"
• OpenAI, "Bots" (crawler documentation for OAI-SearchBot, GPTBot and ChatGPT-User)
• Department of Statistics Malaysia, "Usage of ICT and E-Commerce by Establishment", reference year 2024, released 30 June 2026
• Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan and Ameet Deshpande, "GEO: Generative Engine Optimization", KDD 2024
• Google Business Profile Help, "Tips to improve your local ranking on Google"
• Ipsos, "AI in Asia: Use is running ahead of trust", Ipsos AI Monitor 2026, 22 June 2026
• Acclaira first-party engine probe, 1 August 2026, two queries across ChatGPT, Gemini and Claude with web search enabled and the location set to Malaysia where the engine's API allows it. Per-query answers and cited-source domains retained in the run's probe JSON.
• Acclaira free AEO class (every 2 weeks on a Thursday, 9.30pm Malaysia time, on Zoom, free)