AI Visibility
Year
2026
Focus
AI Visibility
Pipeline Growth
Duration
February – May 31, 2026 (around 15 weeks)
Client
Pharmaceutical companies · Austria
Industry
Pharma / Pharmaceutical Manufacturing, Packaging and Wholesale
Section
Overview
100%
Correct Entity Resolution in 5 AI Systems
43,5%
Brand's AI Visibility Score
116
Total AI Citations
Project Overview
An Austrian pharmaceutical company with a focus on pharmaceuticals/drug manufacturing, packaging, and wholesaling had a problem at the most crucial point in the company: the executive at the top, General Manager and previously active in another role, was still linked by AI systems to their former position and former employer. Anyone who asked ChatGPT, Gemini, or Google AI Mode who this person currently is received outdated answers. For partners, investors, and potential clients who are increasingly researching via AI search, this was a trust issue, not a visibility problem in the classic sense.
The execution
01
Challenge
The executive had formally been the General Manager of the current company for a long time. Nevertheless, AI systems continued to link her to her previous position. Persona entities are particularly conservative for AI systems; Google and LLMs only change established associations when there are clear, repeated, and consistent signals across multiple sources. Without active correction, an outdated association can persist for years.
02
Solution
From mid-February to the end of May 2026, for about 15 weeks, we implemented a targeted entity resolution strategy: structured signals across all relevant channels, consistent role descriptions, and repeated, unambiguous linking of the person and the current company across multiple sources. This was tracked and verified via RankScale against 5 AI systems simultaneously: ChatGPT, Gemini, Google AI Mode, Bing Copilot, and Perplexity.
03
Result
Today, all 5 AI systems tested identify the executive 100% correctly as the General Manager of the current company. The brand itself achieves an AI Visibility Score of 43.5%, well ahead of the most visible competitor at 4.8%.
How We Did It
THE CORE TASK: ENTITY RESOLUTION
The challenge was purely technical, but business-critical. It was about changing an AI perception against an existing, historically grown association. From mid-February to the end of May 2026, about 15 weeks, we explained to the client how conservative Google and AI systems are regarding persona entities, and why it is not enough to build the correction on a single channel.
The test case, ChatGPT asked: "Who is known as the General Manager of [Company]?"
Answer before correction: outdated association with the previous position.
Answer after correction: correct, current role, with a source reference to the current company affiliation.
BRAND PERFORMANCE METRICS
Metric / Value
AI Visibility Score / 43.5%
Sentiment Score / 79.7%
Mentions / 49
Citations / 116
Ø Position / #1.4
Top 3 Visibility / 45.0%
Detection Rate / 42.2%
The most visible competitor was at 4.8% Visibility Score, a significant gap in favor of the brand.
BRAND PERFORMANCE OVER TIME
Brand / 43.5% (+7.8pp)
Competitor A / 7.8%
Competitor B / 4.7%
Competitor C / 4.6%
Competitor D / 4.5%
Competitor E / 4.4%
ENTITY RESOLUTION TEST ACROSS 5 AI SYSTEMS
ChatGPT: correctly identified, current role and company consistently mentioned, old association no longer present.
Google AI Mode: correctly identified, current position confirmed with source references to several independent mentions.
Google Gemini: correctly identified, current position explicitly named as "General Manager and Managing Director", with reference to the relevant business units.
Bing Copilot: correctly identified, including a career history table that clearly shows the current position as dominant, with a high sentiment score.
THE CONCLUSION
When an AI search asks for the decision-maker and provides the wrong answer, it costs trust before a conversation even takes place. This applies to companies just as much as to the people at their helm. This case study shows: persona entity correction is not a side project, but an independent lever for reputation and visibility at the leadership level.
All data captured via RankScale. Tested against ChatGPT, Google Gemini, Google AI Mode, Bing Copilot, and Perplexity.