Google Gemini surpassed Perplexity as the second-largest AI chatbot referral source to websites in May 2026. That’s not a prediction. That’s Statcounter data published last week. ChatGPT still dominates at roughly 60-64% of AI referrals, but the number two spot just changed hands, and most brands have not noticed.
If your AI visibility strategy is “optimize for ChatGPT and Perplexity,” you’re already behind. Here’s what the numbers say, why Gemini’s rise matters, and what you should do about it.
The Data: What Just Happened
Statcounter tracks referral traffic from AI chatbots to websites worldwide. Their latest data shows a clear shift in the competitive landscape:
- ChatGPT: ~60-64% of AI referral traffic (still dominant, but share declining from 70%+ in early 2025)
- Google Gemini: ~15-21% of AI referral traffic (surged from 2.31% to surpass Perplexity)
- Perplexity: growing 370% year-over-year in sessions but now ranked #3 by referral share
- Claude: niche presence with ~14% growth in usage, minimal referral volume
AI search sessions have hit 45 billion per month globally, according to Stackmatix data. This is not a small channel anymore. It’s a significant traffic source that most analytics tools don’t even track correctly. If you’re still measuring success purely through Google Search Console, you’re missing a growing chunk of your actual visibility.
Source: Statcounter Press Release, May 2026
Source: Stackmatix AI Search Market Share Report
Why Gemini Surged: Three Factors
1. Google Distribution Advantage
Gemini is baked into Google Search, Android, Chrome, and Workspace. That’s 4 billion Android devices, 3.5 billion Chrome users, and 2 billion Google Workspace users. When Google integrates an AI assistant into every surface it controls, referral volume follows.
This is not about Gemini being “better” than Perplexity. It’s about distribution. Google doesn’t need users to download an app or change a habit. Gemini shows up where people already are. That’s an unfair advantage that no other AI company can replicate.
2. Google AI Mode Expansion
Google AI Mode is now live in 100+ countries. It’s not just a chatbot. It’s a full search experience powered by Gemini that replaces the traditional blue links with AI-generated answers. BrightEdge data shows AI Overviews appear on roughly 48% of Google queries as of March 2026.
When AI Mode generates an answer that includes a source link, that counts as an AI referral. As AI Mode scales, Gemini referral volume scales with it. This is compounding distribution.
Source: Optimyzee Google AI Overview Ads Data, 2026
3. Perplexity’s Growth Is Real But Different
Perplexity grew 370% year-over-year in total sessions. That’s remarkable. But session growth doesn’t equal referral growth at the same rate. Perplexity is moving toward “answer engines” that synthesize information without always linking out. The more complete their answers become, the fewer clicks they generate.
This is the same pattern Google followed with featured snippets and AI Overviews. Answer quality goes up. Click-through rate goes down. Perplexity’s user base is growing, but each user generates fewer outbound visits over time.
What This Means for Your AI Visibility Strategy
You Need a Multi-Engine Strategy
Optimizing for a single AI engine is like doing SEO for Google in 2005 and ignoring Yahoo, Bing, and Ask Jeeves. The market is fragmented and getting more so. Your brand needs to appear in:
- ChatGPT (still 60%+ of AI referrals)
- Gemini (fastest growing, deepest distribution)
- Perplexity (growing user base, high-intent queries)
- Google AI Overviews (48% of Google queries now show AI answers)
Each engine has different citation patterns. ChatGPT relies heavily on structured content, FAQ schema, and entity authority. Gemini draws from Google’s Knowledge Graph plus web content. Perplexity favors well-sourced, citation-rich content. Your strategy needs to address all three simultaneously.
Google’s Knowledge Graph Just Became More Important
When Gemini answers a question, it doesn’t just scrape the web. It pulls from Google’s Knowledge Graph, which aggregates entity data from Wikipedia, Wikidata, structured markup, and verified sources. If your brand isn’t in the Knowledge Graph, you’re invisible to Gemini.
This means:
- Claim and optimize your Google Business Profile
- Ensure consistent NAP (name, address, phone) data across the web
- Build entity authority through mentions on high-trust domains
- Implement comprehensive JSON-LD structured data on your site
These are not new tactics. But their importance just multiplied because Gemini’s referral share is growing.
Zero-Click Is Accelerating Across All AI Engines
Bain research shows 80% of consumers rely on zero-click results at least 40% of the time. Google AI Overviews cannibalize organic CTR by 42-68% when present. Perplexity synthesizes answers that reduce the need to visit any website.
The implication is stark: getting cited by an AI engine does not guarantee traffic. But it does guarantee visibility. And in a world where AI shapes purchase decisions, brand visibility without clicks still drives revenue. People ask ChatGPT for product recommendations, see your brand mentioned, then go directly to your website or search for you on Google later.
This is the new brand awareness funnel. It doesn’t start with Google. It starts with being the answer AI gives.
Source: Bain Zero-Click Research, 2026
How to Optimize for Gemini Specifically
1. Structured Data Is Non-Negotiable
Gemini processes JSON-LD schema markup directly. Your FAQ schema becomes a citation source. Your organization schema builds entity recognition. Your product schema feeds recommendation engines.
If you’re not using structured data, Gemini has to guess what your content means. When it guesses, your competitors who made it explicit win.
2. Answer-First Content Structure
AI engines extract the first one to two sentences of any section 73% of the time, according to our internal testing at searchless.ai. That means your opening sentence needs to answer the question directly. Background, context, and nuance come after.
Bad: “In today’s rapidly evolving digital landscape, many businesses are wondering about AI visibility…”
Good: “Your brand needs structured data, answer-first content, and entity authority to appear in Gemini results. Here’s how each one works.”
3. Build Entity Authority Across Multiple Domains
The 2026 AI Search Visibility Report found that AI engines cite brands mentioned across six or more high-authority domains. A single blog post on your own site doesn’t build entity authority. You need mentions on Reddit, industry publications, review sites (Trustpilot, G2), and news outlets.
Gemini specifically weighs Google-trusted sources heavily. A mention in a Google News-indexed publication carries more weight than a mention on a random blog.
Source: 2026 AI Search Visibility Report
4. Create an llms.txt File
95% of websites don’t have an llms.txt file. This is the equivalent of not having a robots.txt in 1999. llms.txt gives AI engines a structured summary of your content, products, and brand. Without it, you’re forcing AI engines to infer what you do from random page crawls.
For Gemini specifically, llms.txt complements the Knowledge Graph data. It’s an explicit signal that says: “Here’s what we do. Here’s our content. Cite us accurately.”
5. Monitor Your AI Visibility Across All Engines
You can’t optimize what you don’t measure. Track your brand’s appearance rate across ChatGPT, Gemini, Perplexity, and Google AI Overviews. Track competitors. Track changes over time.
This is exactly what we built at searchless.ai. Our AI visibility scoring checks whether your brand gets recommended across the four major AI engines and gives you a quantifiable score you can improve.
The Competitive Landscape: Where This Is Heading
The AI search market is converging on a model where three or four major engines control how billions of people discover brands, products, and services. ChatGPT has the early lead. Gemini has the distribution advantage. Perplexity has the high-intent user base. Google AI Overviews has the installed base.
Here’s the uncomfortable truth: the brands that win AI visibility in 2026 will compound that advantage for years. AI engines learn from their training data and citation patterns. If ChatGPT starts recommending your brand today, that recommendation reinforces itself in future training cycles. Early movers get a structural advantage that becomes harder to displace over time.
The window is open now. It won’t stay open forever.
FAQ
Is Gemini replacing Perplexity as an AI search engine?
No. Perplexity is still growing rapidly at 370% year-over-year in sessions. But Gemini surpassed Perplexity in referral traffic volume due to Google’s massive distribution advantage. Both matter for your AI visibility strategy.
Should I stop optimizing for Google SEO?
Absolutely not. Google still processes 8.5 billion searches per day. But traditional SEO now competes with AI Overviews, which appear on 48% of queries and reduce organic CTR by 42-68%. You need both traditional SEO and a GEO strategy for AI engines.
How do I check if my brand appears in Gemini results?
Search for your brand name and key product terms in Gemini. Better yet, use a tool that tracks AI visibility across multiple engines simultaneously. At searchless.ai, we offer a free AI visibility score that checks ChatGPT, Gemini, Perplexity, and Google AI in one report.
What’s the difference between optimizing for ChatGPT vs. Gemini?
ChatGPT relies more on training data, web crawls, and structured content like FAQ schema. Gemini draws heavily from Google’s Knowledge Graph, meaning Google Business Profile optimization, structured data, and entity authority matter more. Both reward answer-first content and broad web presence.
How much traffic do AI referrals actually drive?
It varies by industry. Our data at searchless.ai shows AI referral traffic growing 520% year-over-year across all verticals. For B2B SaaS companies, AI referrals now account for 5-15% of total traffic. For consumer brands, the range is 3-8%. The number is growing every quarter.
Does llms.txt actually matter for Gemini?
Yes. llms.txt provides a structured summary that helps all AI engines, including Gemini, understand your content efficiently. Combined with JSON-LD schema and Knowledge Graph presence, it’s part of the three-signal framework that makes AI engines cite you consistently.
What to Do Right Now
- Check your AI visibility across all four engines (ChatGPT, Gemini, Perplexity, Google AI Overviews). You can get a free score at audit.searchless.ai.
- Add structured data (JSON-LD) to every important page on your site. Start with Organization, FAQ, and Product schema.
- Create an llms.txt file if you don’t have one. It takes five minutes and puts you ahead of 95% of websites.
- Restructure your content to be answer-first. The first sentence of every section should directly answer the question the section addresses.
- Build entity authority through mentions on six or more high-trust domains. Reddit, industry publications, and review platforms all count.
The AI search market is shifting fast. Gemini’s rise to number two is a signal, not noise. The brands that respond now will be the ones AI recommends tomorrow.
Get your free AI Visibility Score in 60 seconds at audit.searchless.ai. See what ChatGPT, Gemini, Perplexity, and Google AI actually think of your brand.