For years, companies have treated Google rankings as the front door to online discovery. A new study from Manila-based digital marketing agency Spiralytics suggests that door is no longer the only one that matters.
As consumers increasingly ask ChatGPT, Google Gemini and Perplexity for product recommendations, comparisons and buying advice, brands face a new visibility test: whether artificial intelligence systems mention them at all.
In the Philippines, the early results look encouraging on the surface. Spiralytics found that Philippine brands appeared in 83.2 per cent of the AI answers where they were deemed relevant, compared with 40 per cent for a comparison group of US brands.
Also Read: The Philippines does not need to build AI to have an AI advantage
But the more important finding may be who is doing the talking. According to the study, 95 per cent of the sources cited by AI engines when discussing Philippine products and services came from third parties rather than the brands’ own websites. In practice, that means news articles, online directories, industry roundups and other external references are helping decide how companies appear in AI-generated recommendations.
For Southeast Asian businesses that have spent years optimising websites for search engines, the study points to a shift in how online trust is being assembled. It is no longer enough to publish a clean product page or rank for a keyword. AI systems appear to be piecing together brand identity from a wider set of signals, including whether credible external sources can confirm what the company does.
A new scoreboard for online visibility
Spiralytics conducted the study between 4 June and 3 August 2026. It generated 1,146 real buyer questions for 26 brands, screened the questions for relevance, and evaluated responses across ChatGPT, Google Gemini and Perplexity. In total, the agency reviewed and stored 1,911 AI answers.
The study frames AI visibility as a second scoreboard alongside traditional search engine optimisation (SEO), the practice of improving a website so that it ranks higher on search engines such as Google. AI visibility is slightly different. It asks whether a brand is named, recommended or correctly understood when a consumer turns to an AI assistant for advice.
That distinction matters because Southeast Asia’s internet economy is highly fragmented. Consumers often move between search, social media, marketplaces, chat apps and review sites before making a purchase. In markets such as the Philippines, Indonesia and Vietnam, where mobile-first behaviour is the norm, conversational AI could become another layer in that discovery journey rather than a replacement for search.
The risk for brands is that AI answers can feel authoritative even when they are incomplete. If a company is missing from a relevant recommendation, misidentified, or described through outdated third-party material, the consumer may never visit its website to verify the details.
Third parties hold the pen
One of the study’s sharper findings is that Philippine brands are visible, but not necessarily in control of their visibility. While they appeared frequently in AI answers, their own websites were rarely the cited foundation for those responses.
This should concern marketing and communications teams. Corporate websites remain useful for explaining products, pricing, leadership and services. But if AI engines are leaning heavily on external sources, then earned media, directories, analyst-style lists and independent explainers become more influential in shaping how a company is represented.
“Surprisingly, backlink authority did not predict AI visibility anywhere in our data,” said Gabe Tagala, Head of SEO at Spiralytics. “It does not mean SEO is obsolete. Links still matter where rankings live. But AI keeps a different score: whether the engine can tell which company you are, and if reputable third parties can vouch for you.”
That point is especially relevant in Southeast Asia, where many companies operate through holding structures, subsidiaries, regional brands and local-language identities. A business may be well known offline or on social platforms, but still poorly defined in the open web sources that AI models use to generate answers.
Gemini led on Philippine queries
The study also found meaningful differences between AI engines. For Philippine queries, Google Gemini had the highest brand mention rate at 62.6 per cent, followed by Perplexity at 57.4 per cent and ChatGPT at 45.2 per cent.
Spiralytics described the engines as behaving like “different editors”. That is a useful way to think about the problem. Each tool may weigh sources, freshness, citations and confidence differently. A brand that appears in one engine’s answer may be absent in another.
Also Read: Google bets on schools, telcos and local languages to scale Gemini in Southeast Asia
The study found that absence can also persist. Philippine brands were missing from the three engines’ responses for one in six commercial questions where they should have appeared. In repeat testing on one brand across 11 identical runs, 90 per cent of unanswered questions stayed unanswered each time. Spiralytics said this suggests omission may continue until a brand’s identity signals or third-party coverage change.
For companies, this creates a new monitoring challenge. Tracking search rankings alone will not show whether AI systems are recommending a competitor instead, or whether they are failing to mention the brand in a category where it has a legitimate claim.
The local naming problem
The research also highlights a familiar problem in emerging markets: names do not always travel cleanly across databases and AI systems. Spiralytics found that engines sometimes confused Philippine entities with overseas namesakes, declined to identify a company clearly, or folded subsidiaries into parent conglomerates.
This is not a minor technical issue. In sectors such as fintech, healthcare, education and logistics, a mistaken identity can damage trust or divert demand. For regional companies expanding across Southeast Asia, the challenge compounds as brands adapt to different languages, regulators and market conventions.
The comparison with the US market is revealing. In the US sample, competitor websites accounted for 60.7 per cent of AI citations. In Philippine answers, competitor websites accounted for 36 per cent. That leaves more room for editorial and directory sources to shape AI responses in the Philippines, at least for now.
Jimmy Cassells, co-founder and CEO of Spiralytics, argued that this gap may close quickly. “This window won’t stay open,” he said. “Brands who will move now and treat SERP and AI visibility as a connected strategy are the ones that will still be recommended a year from now.”
Also Read: The AI marketing backlash story doesn’t actually fit Southeast Asia
The broader lesson for founders and operators is simple. AI discovery is becoming part of brand building. Companies need clear websites, but they also need consistent third-party validation and unambiguous public information about who they are, what they sell and where they operate.
For Southeast Asian startups, this is both a threat and an opening. Larger incumbents may have bigger marketing budgets, but younger companies can still earn visibility if they are clearly described across credible public sources. In an AI-mediated internet, the brands that are easiest to understand may become the easiest to recommend.
The post Why Philippine companies are winning AI mentions but losing control of context appeared first on e27.