How visible are Singapore businesses in AI search? On 25 September 2026 the eight Singapore company sites on Google page one for "AI search optimization agency Singapore" served a median 233 KB of HTML. Seven of the eight answered a normal fetch. Two publish FAQPage schema that answer engines can quote directly; two publish no structured data.
The sample is the buyer's own shortlist: the Singapore company websites Google puts on page one for the service this site sells. If you are choosing an agency, these are the pages you would land on first — so it is worth knowing which of them a crawler can actually read and quote.
| Site | HTML served | FAQPage schema | Data blocks |
|---|---|---|---|
| mediaplus.com.sg | 642,128 B | Yes | 7 |
| qoulomb.com | 178,451 B | No | 0 |
| exabytes.sg | 965,373 B | No | 0 |
| purpleclick.com | not measured | not measured | not measured |
| aistudio.com.sg | 86,750 B | No | 4 |
| awebstar.com.sg | 241,972 B | Yes | 2 |
| w360.asia | 333,650 B | No | 1 |
| bthrust.com | 239,024 B | No | 3 |
Median HTML weight: 239,024 bytes (233 KB), range 86,750 to 965,373 bytes. FAQPage schema: 2 of the 7 measurable sites (mediaplus.com.sg, awebstar.com.sg). No structured data at all: 2 of 7 (qoulomb.com, exabytes.sg). AI crawler blocks in robots.txt: 0 of 8 — none of the readable robots.txt files used Disallow: / against GPTBot, PerplexityBot, ClaudeBot, Google-Extended, CCBot or OAI-SearchBot. purpleclick.com answered HTTP 202, a bot-queue response, so its page and robots file could not be measured and it is excluded from every count rather than scored as a failure.
Access is not the problem in this market — nobody in the sample blocks the crawlers. Structure is. Two of the seven measurable sites publish no structured data, so no engine can read a labelled question and answer from them. Only two publish FAQ schema at all. And six of the eight serve between 178 KB and 965 KB of HTML for what is, in most cases, one service page.
That is the opening. An answer engine has to fetch, parse and locate a sentence before it can quote it. A 19 KB page that states a price, a date and a delivery time in plain HTML is easier to quote than a 943 KB page where the same facts sit behind a script.
The part that is not flattering, stated plainly. AIFindr is not in this sample. On 25 September 2026 this site did not appear in the top ten Google results for "AI search optimization agency Singapore", and it did not appear for "AI visibility tool Singapore" either. We are printing that rather than leaving it out. What we can show today is the floor: crawler access, structured data, page weight and quotable passages — all four are on this site now, and they are what we fix on yours.
Method, so you can repeat it. One HTTP GET per URL with a desktop browser user-agent, no cache, 25 September 2026 between 03:10 and 03:15 SGT. HTML weight is the UTF-8 decoded body length of the served page. Data blocks are <script type="application/ld+json"> blocks that parse as JSON. robots.txt is parsed for a Disallow: / rule under a user-agent named GPTBot, PerplexityBot, ClaudeBot, Google-Extended, CCBot or OAI-SearchBot. The sample is the Singapore company websites returned by Google for the query "AI search optimization agency Singapore" (Singapore location, English), excluding the directory listing result. Counts exclude any site we could not fetch; that is why the FAQ figure is out of seven, not eight.
On 25 September 2026 the eight Singapore company sites on Google page one for "AI search optimization agency Singapore" served a median 233 KB of HTML; the heaviest served 943 KB. Seven of the eight answered a normal fetch. Two publish FAQPage schema that answer engines can quote directly; two publish no structured data.
Ask the assistant the question your customer would ask, with your city in it, and read the answer for your business name. Then check the floor an engine needs: robots.txt that does not block GPTBot or PerplexityBot, FAQPage schema, a page under a few hundred kilobytes, and one self-contained 40 to 60 word answer carrying a number.
In our 25 September 2026 sample of eight Singapore sites, none of the robots.txt files we could read blocked GPTBot, PerplexityBot, ClaudeBot, Google-Extended, CCBot or OAI-SearchBot. The gap is not access. It is that two of the seven measurable sites publish no structured data and several serve pages over half a megabyte.
No. On 25 September 2026 aifindr.scalesmes.com did not appear in the top ten Google results for "AI search optimization agency Singapore" or for "AI visibility tool Singapore". We are publishing that number rather than leaving it out, and it is the reason this page focuses on the measurable floor instead of a ranking promise.
Send your website address and we run the same five checks on your live pages plus three competitors, and send back your GAS Score. Free, no card.
Get the free audit See the paid tiersRelated: run the five-minute check yourself · AIFindr pricing · what AIFindr does.