• by Laurie Sullivan  @lauriesullivan, April 9, 2025

BrightEdge launched a feature Wednesday—AI Catalyst—that allows marketers to track and understand how their brands are perceived across generative AI (GAI) search engines during one of the most significant changes to online advertising and marketing.

The feature focuses on presence, perception, and performance by identifying brand sentiments across multiple engines such as ChatGPT, Perplexity and Google’s AI Overviews. The data serves up in a dashboard so marketers can track and measure, negative or positive, whenever information about a company or brand serves up.

It builds on two-decades of historic BrightEdge data. It identifies prompts and keywords that marketers need to monitor across engines. It also displays the data on a dashboard to benchmark results against the competition in real time, and closes the loop to measure, monitor, and perform content over time.

By leveraging billions of historical and real-time data points, marketers can take action on revenue-generating touchpoints, how a brand is perceived, across both traditional search and AI search.

“This isn't just a technological shift,” said Jim Yu, BrightEdge CEO and co-founder. “It is transforming consumer behavior.”

Recent BrightEdge data shows referrals from GAI search engines—ChatGPT, Perplexity, DeepSeek, Gemini, and Claude—rose in February by 29%, 20%, 108%, 16%, and 82%, respectively.

BrightEdge analyzed thousands of brand-specific queries across major AI search engines and found a consistent pattern. Positive mentions only account for one in three mentions in GAI search results, and only one in five of those positive mentions directly recommend a brand.

When AI content is positive—which only accounted for about 31% of the time—40% emphasize quality and capabilities, 23% describe the brand as exceptional, and 21% directly recommend the brand.

For neutral mentions, which made up 66% of all brand mentions, AI typically provided straightforward information, how-to content, or technical specifications without any evaluative language.

Though negative mentions only account for 3%, when they occur, 40% highlight product limitations, 30% discuss limitations for specific features, and 20% cite specific user issues.

“This means that even if your brand is found, it's being presented and perceived differently depending on which AI engine a person uses,” according to the data.

Yu called this "one of the most significant inflection points for brands in history because if brands do not inject themselves into these key AI conversational moments, AI will do it for them."

When a consumer asks Perplexity to cite the most reliable SUVs, "it doesn't just return links—it crafts a story. It decides which brands get mentioned, which models are recommended, and how they're positioned," Yu explains.

When a company shops for human resource software, the AI doesn't just list names. It sets the criteria, decides what matters, and builds a shortlist in seconds.