
May 26, 2026 · 8 min
A SearchBrand.ai study analyzing more than 1,500 queries on ChatGPT, Gemini, Claude and Perplexity reveals that 73% of Chilean consumers already prefer using AI over traditional search engines when researching products online. The problem: 67% of brands have no presence in these responses. Only 14% of companies account for 78% of recommendations. Days before CyberDay 2026, the gap between visible and invisible brands in AI is defining who sells and who disappears.
Five days before CyberDay 2026 (scheduled for June 1–3, according to the Santiago Chamber of Commerce), a silent change already redefined how Chileans decide what to buy. The traditional search engine, with its ten blue links and paid ads, is losing ground to a new interface: conversation with an AI assistant.
The “Digital Chilean Consumer X-Ray”, developed by SearchBrand.ai, evaluated more than 1,500 purchase intent queries across four AI platforms: ChatGPT, Gemini, Claude and Perplexity AI. Categories analyzed were retail, technology, travel and apparel — the four most in-demand verticals during CyberDay.
The main finding: 73% of surveyed consumers declared that they prefer a direct answer from an AI assistant rather than navigate multiple websites. This is consistent with an Ipsos and Google study from March 2026 showing 66% of Chileans still use Google Search as a starting point, but now complemented by an AI layer that modifies the decision flow.
The first finding is the visibility gap: 67% of the analyzed brands rarely or never appear in AI assistant recommendations for offers, product comparisons or convenient brands for CyberDay.
The second is concentration. 78% of the total recommendations are concentrated in just 14% of the analyzed companies. Brands with technical optimization or consolidated topical authority capture almost all mentions.
Among younger consumers (18–35), 61% already use AI as their first stop to compare prices and ask for recommendations. And 58% admitted changing their purchase decision after interacting with an AI assistant.
Hussam Sufan, CEO and founder of SearchBrand.ai, summarizes it: “Consumers no longer just enter keywords. Now they converse, argue and ask for personalized recommendations. The purchasing decision is shifting to AI chats.”
The Santiago Chamber of Commerce turned it into official infrastructure. On May 26, 2026, the CCS announced “Cyber AI”, a generative AI conversational agent that will act as a 24/7 advisor on cyber.cl.
The CCS projected USD 540 million in sales for this CyberDay, 5% growth. Chile's e-commerce is expected to reach USD 9,000 million in 2026, at 15% retail penetration. How many sales will pass through an AI filter before being completed?
Felipe Almeida, co-founder of SearchBrand.ai, points to a technical cause: “Many companies still don't understand that if their websites aren't optimized for AI crawlers to read and understand them, they start to disappear from the customer's mental purchase process.”
Language models don't browse the web like humans. They don't click buttons, execute JavaScript or interpret product images. They read plain text, metadata and structured data. An e-commerce with image-based product pages and no complete textual descriptions is, for an AI, a blank page.
The second layer is the absence of topical authority. A Diario Estrategia study confirms 47% of consumers turn to AI for research before buying, and 68% of frequent users trust the information they receive.
A concrete example: Tenpo, the Chilean fintech, worked with SearchBrand.ai to measure and improve its AI presence. After implementing technical optimizations (structured data, conversational content, natural-language FAQs), Tenpo went from appearing in 2 of 25 relevant questions to being mentioned in 15 of 25.
At the CCS seminar on AI and GEO (May 14, 2026), specialists recommended four concrete actions:
During eCommerce Day Chile 2026 (April 9), experts detailed that AI-assisted e-commerce queries grew 551% during Look Week 2026. Growth is exponential; late-arriving brands will compete against already consolidated positions.
A NubeCommerce 2026 report confirmed 56% of Chilean online stores already operate with some AI component. And 17.5% anticipate that autonomous AI agents will drive the most impactful commerce technology of the year.
The SearchBrand.ai platform lets brands measure this visibility in real time. Its AI Scoring returns a 0–100 index comparable across time and brands.
The message for the 572 CyberDay participants is clear: they will compete not only on price and discount, but on appearing in the answer when a consumer asks ChatGPT: “Where can I find the best CyberDay deals?”
Based on the “Digital Chilean Consumer X-Ray” by SearchBrand.ai, analyzing 1,500+ simulated purchase intent queries across ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic) and Perplexity AI in retail, technology, travel and apparel. Data was complemented with public sources from the Santiago Chamber of Commerce, eCommerce Day Chile 2026 and industry reports.
According to a SearchBrand.ai study that analyzed more than 1,500 queries on ChatGPT, Gemini, Claude, and Perplexity, 73% of Chilean consumers prefer to use AI assistants rather than browse multiple websites when looking for products and online offers.
Only 14% of the analyzed companies concentrate 78% of the recommendations generated by AI assistants. 67% of the evaluated brands have low or no presence in the responses of ChatGPT, Gemini, and other AI platforms.
Yes. 58% of consumers recognize having changed their purchase decision after interacting with AI recommendations. This percentage is higher among users aged 18 to 35, where 61% use AI as their first option to compare prices and ask for recommendations.
AEO (Answer Engine Optimization) is the discipline that optimizes a brand's presence in the responses of AI assistants like ChatGPT, Gemini, and Perplexity. Unlike traditional SEO, AEO focuses on getting AI to mention and recommend the brand when a consumer asks conversational questions.
By implementing JSON-LD structured data on their sites, creating content in a question-and-answer format with natural language, publishing an llms.txt file in the root directory, and ensuring product pages have complete textual descriptions that language models can read.