Philipp Kaufmann
Research Associate & Doctoral Candidate
University of St.Gallen
Philipp Kaufmann is a research associate and doctoral candidate at the Institute for Marketing and Customer Insight of the University of St.Gallen in St. Gallen, Switzerland. A former digital marketer at the Swiss comparison platform Comparis, he co-wrote a 2026 study showing that brand visibility in AI search shifts so much from day to day that a single check is unreliable.
Why One AI Visibility Check Is Not Enough
"In short, AI visibility should be managed as a probability over repeated measurements, not as a one-time ranking snapshot," Kaufmann wrote on LinkedIn in April 2026, when he and his St.Gallen colleagues Julius Schulte and Malte Bleeker published "Don't Measure Once: Measuring Visibility in AI Search (GEO)."
The team tracked four AI search engines, including ChatGPT, Gemini, and Google's AI Mode, across four Swiss-German campaign categories from January 24 to March 20, 2026, using prompts drawn from high-volume search queries. Only 34% to 42% of cited sources carried over from one day to the next, meaning roughly 65% of sources turned over daily. Brand mentions were steadier, with 45% to 59% overlap, but varied widely. Even identical prompts run at the same moment produced different sources and brands.
A Measurement Playbook for Marketers
The authors argue that AI search works on inclusion and exclusion rather than ranking. In classic SEO a page may slide down the results, while in an AI answer a brand can appear in one response and vanish from the next. Their paper recommends at least seven runs per prompt per day to estimate brand visibility and eight when cited sources matter, along with a broad set of prompts and observation windows of several weeks rather than days.
"Measuring a brand's AI visibility is becoming increasingly important as users' search behaviour shifts from traditional search towards AI search," Kaufmann wrote. The study also found that a handful of domains captured most citations in every campaign, with Google's AI Mode the most concentrated of the four engines, and Kaufmann recommends setting separate benchmarks for each AI platform.
From Comparis to St. Gallen
Kaufmann spent six years in digital marketing at Comparis, the Swiss comparison platform, before joining the University of St.Gallen in November 2023. He works as a doctoral student and research assistant for Sven Reinecke at the Institute for Marketing and Customer Insight, where he also leads the editorial team of the practitioner journal Marketing Review St.Gallen. His research covers marketing management and performance marketing, with a focus on digital and product marketing and web analytics.
He holds a bachelor's degree in business administration from the Lucerne University of Applied Sciences and Arts in Lucerne, Switzerland, and a master's degree in innovation management from the University of Manchester in Manchester, England.