AI Brand Visibility: We Tested 5 Brands Across ChatGPT, Claude, and Gemini

Honey was acquired by PayPal for $4 billion. Ask ChatGPT “What is Honey?” and it tells you about bees. This isn’t a bug. It’s a AI brand recognition problem that affects thousands of brands, and most don’t even know it exists.

As AI assistants become the default way people search for information, a new challenge is emerging: brand visibility in AI search.

When someone asks Google “What is Copper?”, the search engine returns a mix of results, including Copper CRM. The user clicks, browses, and decides.

When someone asks ChatGPT the same question, they get a single answer. And if that answer is about the metal instead of the CRM software, the brand never had a chance.

We call this the Context Gap: the difference between how AI responds when given context (”What is Copper, the CRM?”) versus when asked directly (”What is Copper?”).

Users don’t naturally provide context. They ask simple questions. And simple questions are where brands disappear from AI search results.

We tested 5 well-funded tech brands with common English word names:

• Honey: Coupon browser extension, acquired for $4B by PayPal

• Heap: Product analytics platform, $110M+ raised

• Pitch: Presentation software, $100M+ raised

• Copper: CRM for Google Workspace, $90M+ raised

• Grain: Meeting recording tool, $20M+ raised

We tested each brand with 3 prompt variants:

• With Category: “What is [Brand], the [category]?” (e.g., “What is Copper, the CRM software?”)

• With Competitor: “Compare [Brand] vs [Competitor]” (e.g., “Compare Copper vs Streak”)

• Alone: “What is [Brand]?” (no context)

Each variant was tested across 3 major AI models: ChatGPT (GPT-5.2), Claude (Sonnet 4.5), and Gemini (3 Flash).

100% means the AI correctly recognized the brand/product
50% means the AI asked for clarification (”Which Copper do you mean?”)
0% means the AI confidently gave the wrong answer (explained the common word)

When users ask about a brand without context, AI brand recognition drops dramatically:

• Honey: 0% on ChatGPT and Claude (AI explains bee honey)

• Copper: 0% on ChatGPT and Claude (AI explains the metal)

• Pitch: 0% on ChatGPT (AI explains sound frequency)

These aren’t small startups. They’ve raised hundreds of millions of dollars combined. Yet ChatGPT doesn’t mention them when users search.

With category context, recognition jumps to 100% on ChatGPT and Claude. The AI knows exactly what you’re asking about.

The problem? Users don’t naturally include context. They type “What is Honey?” not “What is Honey, the browser extension?”. This is the core AI search optimisation challenge.

ChatGPT and Claude show nearly identical behavior: excellent AI brand recognition with context, poor without. Gemini 3 is more cautious, asking for clarification more often (resulting in more 50% scores), but this means users still don’t get a direct answer about the brand.

Adding a competitor doesn’t always help. When both the brand and competitor have common word names, confusion persists:

• “Compare Copper vs Streak” resulted in ChatGPT asking for clarification (50%)

• “Compare Grain vs Otter” had mixed results across models

As traffic shifts from traditional Google search to AI assistants, brand visibility in AI responses becomes critical. This isn’t about ranking on page 1 anymore. It’s about being the answer when ChatGPT, Claude, or Gemini responds to a user query.

This shift is sometimes called Generative Engine Optimization (GEO) or AI Search Optimization. Whatever you call it, the goal is the same: ensuring AI assistants recognize and recommend your brand.

Every time AI fails to recognize your brand:

• A potential customer gets the wrong information

• Your competitor might get mentioned instead

• The user never discovers your product existed

• Monitor your AI brand visibility across different prompt types and models

• Identify context gaps where your brand fails without explicit context

• Optimize your digital presence to strengthen brand signals that AI models can learn from

• Track changes over time as models are updated and retrained

The brands that figure out AI search optimization first will have a significant advantage as more users shift from Google to AI assistants for product research and recommendations.

• Models tested: ChatGPT (GPT-5.2), Claude (Sonnet 4.5), Gemini (3 Flash)

• Test date: January 2026

• Sample size: 5 brands, 3 variants each, 3 models = 45 data points

• Limitations: Small sample size, English-language prompts only

This study is directional, not statistically significant. Results may vary based on prompt phrasing, model updates, and other factors.

We’ll test your brand across multiple AI models and prompt types, showing you exactly where you’re visible and where you’re invisible in AI search.

This research was conducted by friction AI, a platform that helps brands monitor and improve their visibility across AI search platforms, including ChatGPT, Claude, and Gemini.

Run your free AI visibility report here.