BrandSource.AI Research & Insights

Research and analysis on how AI models learn about brands and how to improve AI accuracy.

• How Category Context Shapes What AI Says About Your Brand — and How to Control It — AI models don't describe brands in isolation. They describe them through the lens of whatever category they've been filed in. Getting categorized wrong is one of the most persistent and damaging AI brand problems.
• The Rebranding Problem: How AI Systems Handle Company Name Changes — and Why They Fail — Rebrands, acquisitions, and pivots are some of the most common sources of AI brand hallucinations. The model learned your old identity and the update hasn't propagated. Here's what's actually happening and how to manage it.
• Why Your Founder's Personal Brand Is a Trust Signal for AI Systems — AI systems don't just learn about companies — they learn about the people associated with them. A founder with a documented public presence dramatically improves AI recall accuracy for their company. Here's how and why.
• The Cold Start Problem: How New Brands Can Establish AI Identity Before They Have History — New brands face a catch-22 in AI: AI systems learn from web history, and new brands don't have any. Here's how to solve the cold start problem and build AI brand identity from day one.
• Why JSON-LD Is the Highest-Signal Format for AI Crawlers — and How to Use It Correctly — Of all the content formats AI crawlers process, JSON-LD structured data consistently produces the most reliable extraction results. Here's the technical why, the common mistakes, and a complete Organization schema template.
• How to Audit Your Brand's AI Representation: A Practical Step-by-Step Guide — Most brands have no idea what AI systems are saying about them right now. This is a complete, reproducible process for auditing your brand's AI representation — what to test, how to score it, and what to do with what you find.
• Why Small Brands Are Disproportionately Hallucinated by AI — and What to Do About It — The smaller your brand, the more likely AI is to get you wrong, confuse you with a competitor, or invent facts about you. Here's the data on the small brand hallucination gap — and the specific steps to close it.
• The Brand Identity Stack: Seven Layers That Determine AI Recall Quality — AI brand recall isn't a single thing — it's the result of a layered information architecture. Brands that understand all seven layers can systematically close the gaps that cause hallucinations.
• Perplexity vs. ChatGPT: Why Your Brand Strategy Needs to Address Both — Retrieval-augmented AI and training-based LLMs represent two fundamentally different ways your brand gets described by AI — and they require different interventions. Most brands are only addressing one.
• What Actually Happens to Your Brand Data Between Crawl and Model Training — AI crawlers visit your site, but that doesn't mean tomorrow's ChatGPT knows your brand. The pipeline between crawl and model recall is longer, lossier, and stranger than most brand teams realize.
• The Future of Brand Identity in the Age of AI — For thirty years, brand identity meant controlling what people saw. In the age of AI, it means controlling what machines know. This is a fundamental shift — and most brands aren't ready for it.
• Why Your Wikipedia Page Isn't Enough for AI Brand Accuracy — Wikipedia is a major source for AI training data, and many brands rely on it as their primary canonical reference. But our research shows it's not sufficient — and sometimes it actively works against brand accuracy.
• AI Visibility Score: How We Measure a Brand's Presence in AI Systems — BrandSource.AI's AI Visibility Score is the first quantitative measure of how well a brand is represented across AI systems. Here's exactly how it's calculated, what it means, and how to improve yours.
• Per-Crawler HTML Variants: How Different AI Bots Prefer Different Content — GPTBot, ClaudeBot, and PerplexityBot don't read content the same way. BrandSource.AI serves different HTML variants to each — and the early results of this A/B experiment are reshaping how we think about AI content optimization.
• AI Brand Hallucinations: What They Are and How to Fix Them — AI models confidently state wrong facts about brands every day — wrong founding dates, discontinued products, merged companies. This is the definitive guide to understanding why it happens and what you can do about it.
• How AI Models Learn What They Know About Your Brand — We analyzed 100,000+ AI crawler visits to BrandSource.AI and discovered a clear pattern: the structure and format of brand information dramatically affects whether AI models recall it accurately. Here's what we found.