Peec AI assembles research team to decode AI brand recommendations

Peec AI has hired a research team to dig into why ChatGPT, Perplexity and Google’s AI Mode recommend one brand instead of another. The June 26 move pairs former enterprise SEO leaders, an explainable-AI researcher and engineers behind some of the market’s earliest tools, with a clear goal: reverse engineer the logic behind AI search and push the findings back into the product.

Peec AI already sells itself as an AI search analytics platform for marketing teams trying to measure brand performance across ChatGPT, Perplexity and Gemini. Its core metrics are visibility, position and sentiment, a framing that goes beyond simple mention tracking and into comparative brand placement. The company has also named Viktoria Bykanova, Lauryn Chamberlain, Tomasz Rudzki, Noah Wolff, Metehan Yesilyurt and Jan Ehrlinspiel in the research push, signaling that Peec AI wants more than dashboards. It wants a working model of why one source wins a recommendation and another gets skipped.

That ambition lands at the exact moment AI shopping and discovery features are becoming more commercial. OpenAI launched shopping research in ChatGPT in 2025 for logged-in users on Free, Go, Plus and Pro plans, then expanded it in March 2026 with richer product discovery, comparisons and merchant integration powered by the Agentic Commerce Protocol. OpenAI says product results are selected independently and are not ads. Google has pushed in the same direction, saying at I/O 2026 that Gemini 3.5 Flash became the default model in AI Mode globally, while its shopping materials now describe visuals, smart guidance, reliable product data and agentic checkout inside AI Mode.

That shift has made AI visibility a more technical game. Yext said it analyzed more than 6.8 million citations across Gemini, ChatGPT and Perplexity and found the models often disagree, which is exactly the kind of inconsistency Peec AI is trying to exploit as a research problem. Peec AI has been publishing its own large-scale data work too, including claims of more than 65 million citations across 20 countries and a separate sample of 500,000 prompts in which AI Overviews appeared 86% of the time.

The company has grown fast enough to make that research posture look less like theory and more like a commercial wedge. In May 2026, Peec AI said it had passed $10 million in annual recurring revenue 16 months after launch, with more than 2,500 customers globally, 1,300-plus brands and agencies onboarded since February 2025, and 300-plus new customers a month. It also opened its first U.S. office in New York, where it says demand is strongest. Founded in 2025 by Marius Meiners and Daniel Drabo, and backed by Singular, Antler, Combination VC, identity.vc and S20, the Berlin company is trying to become one of the standard-bearers in a category that is still deciding how to measure itself.