LLM Monitoring in Action: How an International B2B Trading Company Uses Zeno Visibility to Monitor and Manage Its Brand Presence Across Five AI Platforms Simultaneously
LLM Monitoring in Action How an…
---
Starting Point
MedTrade International is a mid-sized B2B trading company headquartered in Vienna, distributing medical consumables and laboratory equipment to hospitals, clinics, and research institutions across 14 European countries. With annual revenues of approximately €87 million and a product portfolio of over 4,200 SKUs, the company is a well-established supplier in the DACH region and Eastern Europe.
Since 2023, the marketing team had been observing a structural shift in how their target audience approached procurement: purchasing decision-makers at clinics and laboratories were increasingly turning to AI-powered systems — particularly ChatGPT, Perplexity, and Microsoft Copilot — to research suppliers, compare product categories, and build initial shortlists. Internal surveys revealed that around 38 percent of new contacts in 2024 reported having initiated their first interaction with MedTrade based on an AI recommendation — even though the company had never actively engaged with this channel or systematically tracked it.
The core problem: MedTrade had no visibility into whether, how, or in what context AI models were mentioning the company — and had no infrastructure in place to change that.
---
Challenge
The marketing team, led by CMO Sandra Veit, faced a three-dimensional visibility problem. First, there was a complete absence of data: no systematic tracking of whether MedTrade appeared in AI-generated responses at all, in what context mentions occurred, or how the company was positioned relative to competitors such as Medline Europe or Lohmann & Rauscher.
Second, the existing content infrastructure was not built for machine readability: product pages lacked structured markup, internal linking followed no semantic logic, and topical authority had not been systematically developed for a single core subject area.
Third, the team lacked the organizational capacity to act: they had neither the tools nor the processes to monitor five AI platforms in parallel, consolidate findings, and translate them into operational content measures. The risk was tangible: competitors who established themselves earlier as reference providers within AI systems would build structural recommendation advantages that would be difficult to reverse.
---
Solution
Following a three-month evaluation phase, MedTrade International chose to deploy Zeno Visibility — with the goal of building a complete AI Visibility infrastructure that integrates monitoring and content production within a closed-loop system.
Phase 1 — Baseline Measurement (Weeks 1–3): Zeno Visibility's research engine was configured for 22 strategic keywords, including terms such as "medical consumables supplier Europe," "laboratory supplies B2B procurement," and "clinical supplies wholesale DACH." The platform sent structured queries in parallel to ChatGPT (GPT-4o), Gemini 1.5 Pro, Perplexity, Claude 3.5 Sonnet, and Microsoft Copilot, systematically capturing whether and how MedTrade appeared in the responses. The result: an initial Semantic Authority Score of 14 out of 100 — with mentions in only 3 of 22 keyword clusters, exclusively on Perplexity.
Phase 2 — Authority System Build (Weeks 4–12): For each of the six prioritized keyword clusters, Zeno Visibility's Authority System Builder generated a complete content system: hub pages, thematically linked blog articles, FAQ pages, comparison pages, and structured product category descriptions — 134 pieces of content in total, with automatically generated Schema.org JSON-LD markup and a defined internal linking architecture. The content was published directly into MedTrade's existing WordPress CMS, supplemented by a structured export to Contentful for the English-language market presence.
Phase 3 — Continuous Monitoring (from Week 13 onward): Weekly monitoring across all five AI platforms was established as a standard process. Changes in the Semantic Authority Score, new mention contexts, and competitor positioning were automatically captured and delivered to the marketing team as a consolidated report.
---
Results
Six months after project launch, reliable comparative data was available.
The Semantic Authority Score rose from 14 to 61 out of 100 — an increase of 336 percent. MedTrade was now mentioned in 17 of 22 keyword clusters on at least one AI platform, and in 9 clusters on three or more platforms simultaneously.
The platform distribution shifted fundamentally: whereas only Perplexity had generated mentions at the outset, MedTrade was present across all five monitored platforms after six months — with the strongest gains on ChatGPT (from 0 to 11 relevant keyword clusters) and Microsoft Copilot (from 0 to 7 clusters).
In the sales context, the sales team reported a measurable shift in initial conversations: the share of inbound inquiries where prospects explicitly referenced an AI recommendation rose from 38 to 54 percent. Average lead qualification time dropped by 22 percent, as prospects were entering conversations with concrete product knowledge already in hand.
The content output of 134 published pieces generated an additional 2,840 organic sessions per month within six months — a side effect that had not been defined as a primary objective in the original project scope.
---
Lessons Learned
1. AI visibility is not self-regulating. Companies that rank well in traditional search engines are not automatically treated as authorities by LLMs. Semantic interconnection and machine-readable structures must be actively built.
2. Monitoring without the ability to act is worthless. The combination of measurement and autonomous content generation was the decisive factor. Standalone monitoring tools would have made the problem visible — but not solved it.
3. Platform diversity is not optional — it's essential. Different LLMs weight sources differently. A presence concentrated on a single platform is structurally fragile, as MedTrade's initial situation on Perplexity demonstrated.
4. Schema.org markup is not a technical detail. Structured data was a direct lever for inclusion in AI-generated responses throughout this project. Content without machine-readable markup was consistently cited less often.
5. Building semantic authority requires depth, not breadth. Six fully developed keyword clusters far outperformed a surface-level coverage of 22 topics.
---
Summary
Within six months, MedTrade International built a complete AI Visibility infrastructure that transformed the company from a single-platform AI presence into a measurable authority across five LLMs. Zeno Visibility enabled not only the measurement of the status quo, but the autonomous construction of the semantic foundation that AI models require in order to make recommendations. The project demonstrates a clear principle: organizations that treat AI visibility as an infrastructure challenge — rather than a reporting discipline — create structural competitive advantages that translate directly into sales outcomes.
---
More Case Studies
---
*This content was created with AI assistance and editorially reviewed.*