Webull Singapore AEO Assessment Report
by Novastacks AI

The six query sample above is diagnostic. To size the problem properly we ran 46 prompts across Google AI Overview, ChatGPT and Gemini on 11 August 2026, covering the full journey from market discovery through evaluation to account opening. The six original queries are included unchanged so the earlier numbers stay comparable.

The result is unambiguous. Webull is cited when a buyer already knows the name, and effectively absent the moment they describe what they want instead.

The headline 31.1% is carried entirely by branded prompts. Strip the brand name out and citation falls to 14.3%. That is the number that matches the actual commercial problem, and it is the one worth tracking month to month.

The most winnable segment sits between those two columns. On 8 non branded prompts the AI names Webull but sources somebody else's page. The model already wants to talk about Webull and has no Webull page it can use. That is a supply problem, not a reputation problem, and supply is fixable.

Section level citation counts show where the supply already exists and where it is mechanically blocked.

Method note: 868 total citations, of which 526 are real publisher citations after excluding Gemini's redirect wrapper. Two of 46 AI Overview calls returned API errors and are excluded from all rates.

The good news first. Every crawled page on webull.com.sg is server-rendered and fully readable by bots, with no client-side rendering and no empty shells. The crawler rules are a different story. robots.txt carries a User-agent: * group containing Disallow: /quote/ and Disallow: /news/. Googlebot, bingbot and Yahoo Slurp each have their own named groups, so they ignore that wildcard group and crawl freely, which is why Google coverage looks healthy. Every AI crawler without a named group, including GPTBot, ClaudeBot, PerplexityBot, CCBot and Google-Extended, falls under the wildcard and is barred from every quote page and every news page. Our 46 prompt run confirms the effect rather than assuming it: across 868 citations, exactly zero originate from /quote/ or /ticker/. Speed is not the problem either: the homepage records FCP 1.1 s, LCP 2.8 s, CLS 0.009 and total blocking time 208 ms, which beats Tiger Brokers on all four measures and matches moomoo on overall performance.

The problem is discovery. The site's address list at /sitemap.xml redirects to a 404 page and contains zero entries, and the pointer to it inside the crawler-rules file is present but empty, the line reads "Sitemap:" with nothing after it. Both competitors publish working address lists. Combined with the blog index linking only 3 of its 16 articles, 13 long-form guides have no discovery path at all, no menu link, no index link, no address-list entry. Google found five of them anyway and quoted them in AI Overviews. That is the measure of what is being left on the table: the content earns citations on merit while the site withholds every signal that would help it be found.

The overall search-health score reflects it, 77 for Webull against 85 for moomoo and 92 for Tiger Brokers, despite Webull having the fastest-loading homepage of the three on three of four speed measures.

Machines read a page through its heading structure. On webull.com.sg, five of the six crawled page types return zero H2 subheadings, and three of them return no H1 either, the help centre, the education hub and the blog index all open with no top-level heading. The US stocks page runs the opposite error, stacking three competing H1s on one page. moomoo's US stocks page carries 17 subheadings and Tiger's homepage carries 8; Webull's equivalents carry none.

Two further signals compound it. Four of six pages have no canonical address, and the ticker page's canonical points to webull.com, the global domain, not the Singapore one, which tells search engines the Singapore version is a duplicate of a page on another site. And none of the six pages carries any language or region targeting, where moomoo ships 16 to 31 declarations per page and Tiger Brokers 9 to 14. In a market where the same brand operates a US, HK and Australian estate, that is the signal that tells Google which audience each page is for.

The homepage also has no navigation, main or section landmarks in its markup, footer only, which removes the structural cues a crawler uses to separate primary content from furniture.

Structured data is how a brand tells search engines and language models what it is, what it charges and what it answers, without those systems having to infer it from prose. Across the homepage, ticker, US stocks, pricing, education and help pages, webull.com.sg carries none of it. The only markup anywhere on the domain is a BlogPosting block on the 16 blog articles, and that block contains just two fields, a headline and a publish date. No author, no publisher, no image, no update date.

The contrast with both competitors is total. moomoo's homepage declares itself as an organisation with a physical location, an aggregate rating and a full question-and-answer block. Tiger Brokers' homepage declares 18 distinct types, including a financial-services identity, a site-search entry point, an offer, a step-by-step guide and an aggregate rating. Those are the machine-readable facts an AI answer is assembled from, and on the three queries where AI Overviews named Webull but cited someone else, someone else had them.

We ran the three keywords that define this category in Singapore through live search. webull.com.sg does not appear in the top 20 for any of them. All three trigger a Google AI Overview, so the answer is being written before a buyer reaches the results at all.

The most consequential is the middle row. "US stock trading platform singapore" is a near-verbatim restatement of Webull's own homepage headline, "Unlimited Free Trading for US Stocks & ETFs", and five rival brokerages rank their own domains on it: Interactive Brokers Singapore at #3, Standard Chartered at #5, Charles Schwab Singapore at #8, moomoo at #10 and POEMS at #12. Webull's endorsement for exactly this use case exists, Beansprout names it best for "US stocks and ETFs" in the table Google lifts into its People Also Ask box, but the endorsement lives on Beansprout's domain while Webull's own page is nowhere in the results.

Treat this as a signal test, not a census: three keywords and six AI queries are a diagnostic sample. But the sample is unanimous, and it points at territory rather than technique.

Across the 150 highest-traffic keywords each domain ranks for in Singapore, 58.5% of Webull's estimated traffic comes from searches that already contain the brand name. For moomoo the equivalent figure is 1.8%; for Tiger Brokers, 12.5%.

The ranking profile says the same thing in a different way. Webull holds 13 first-place rankings in Singapore, and every one of the 12 visible in the traffic sample is a branded search: "webull", "webull singapore", "webull promotion", "webull promotions", "webull referral", "webull login", "webull desktop", "webull paper trading". These are searches by people who have already decided to look you up. They are worth holding, and they measure existing demand rather than creating new demand.

Where non-branded traffic does arrive, it arrives on ticker pages that rank in the 20s to 40s: "singapore airline share price" at #26, "sgx:o39" at #24, "grab share price" at #46, "nasdaq qqq" at #45. Those are terms carrying 33,100 to 110,000 monthly searches, landing outside the first two pages. Tiger Brokers' two strongest ticker pages alone carry 48,627 of its top-150 traffic; moomoo's single strongest carries 166,791.

moomoo ranks for 13,481 keywords in Singapore against Webull's 311, a 43x gap. Tiger Brokers holds 3,416, or 11x. The traffic gap is wider still: 544,325 estimated monthly visits against 3,681.

Two structural decisions drive most of it. First, both competitors run Singapore as a directory on a global domain, so every Singapore page inherits whole-domain authority and every global surface, quotes, news, community, screeners, counts toward the Singapore footprint. Webull Singapore runs a standalone Singapore domain that starts from zero. Second, both competitors have built page types Webull has not: moomoo's community forums produce pages that rank independently on each ticker, and its news and video hubs publish continuously.

The same ordering shows up in AI answers. Across the Singapore AI-answer corpus, webull.com.sg is cited 33 times, itiger.com 300 times and moomoo.com 1,825 times. The search moat and the citation moat are the same moat.

Four things in this dataset are genuinely strong, and all four are currently working without any structural help.

The blog earns AI citations on merit. Google's AI Overview quoted five different Webull articles during this assessment, /blog/4, /blog/7, /blog/9, /blog/13 and /blog/14, on branded and beginner-brokerage questions. Those pages have no menu link, no index link and no address-list entry. They are being found and quoted anyway.

Help answers already rank as answers. /help/faq/1075 sits at #7 for "what time us market open in singapore"; /help/faq/1361 sits at #13 for "morningstar" (2,900 monthly searches). Question pages winning question searches without question formatting.

Branded search is fully owned. #1 for "webull" (4,400 monthly searches) and "webull singapore" (1,000), plus promotions, referral, login, desktop and paper trading. On the branded AI query, Webull held organic #4, #5, #8 and #9 simultaneously, the entire answer block.

Independent endorsement already exists. Beansprout names Webull Singapore best for "US stocks and ETFs" in the table Google lifts into People Also Ask. MissLobang's article, cited twice by ChatGPT, carries Webull in its headline. Seedly holds 4.1 out of 5 across 26 reviews. Third parties are already saying the right thing; the brand is not yet positioned to be the source they point back to.