These case studies document measurable generative engine optimization results delivered by Latent Analytics: baseline Share of Model, the technical interventions deployed, and the change in LLM citations across ChatGPT, Gemini and Perplexity. Each study includes methodology and timeframe so the results can be evaluated rigorously.
The Challenge: RR Service, an international educational consultancy (“Studia in Europa”), was effectively invisible to generative search. On a fixed set of 5 strategic prompts covering its highest-value programs (medicine and dentistry in Italy, Malta and Portugal), the brand recorded 0 mentions out of 15 observations across ChatGPT, Gemini and Perplexity (May 2026 baseline). The “RR Service” entity also suffered from severe ambiguity: AI systems conflated it with unrelated businesses sharing the same name. The client required a rapid increase in generative visibility without disrupting their existing WordPress infrastructure.
The Architecture: Rather than rebuilding the site, we engineered a modular “Additive Merge” strategy. We developed highly optimized semantic blocks — entity disambiguation, structured key facts, and machine-readable FAQs — injected via shortcodes into their highest-traffic existing URLs. A dedicated Knowledge Block was built to semantically bind the RR Service entity to its core academic partners, such as Universidade Fernando Pessoa (Porto).
The Execution: We deployed the on-page semantic restructuring across the priority URLs and validated machine readability through retrieval simulations, confirming that the optimized pages now surface in the search layer used by AI engines. Phase two — building third-party trust signals through reviews, community presence and earned media — is currently in progress.
The Output (measured, June 2026): One month after deployment, the same 5-prompt protocol — extended to four engines with the addition of Claude — returned 8 mentions out of 20 observations (40%), with the brand surfacing consistently on the same 2 prompts across all four engines (ChatGPT, Gemini, Perplexity and Claude), and 3 rrservice.info URLs now retrievable for the improved prompts. A transition from complete latent invisibility to active generative citation on the brand’s core commercial queries.
The Challenge: Città per te, an established Italian travel agency specializing in curated group tours and extended European itineraries, faced a critical visibility gap. As travelers shifted from traditional search engines to using Large Language Models (LLMs) to plan trips, the agency’s complex catalog of dates, pricing, and destinations was unreadable to generative AI systems.
The Architecture: We conducted a rigorous RAG (Retrieval-Augmented Generation) Audit to analyze how global AI models failed to retrieve their tour data. We then engineered a custom semantic data layer, effectively converting their unstructured web pages and itinerary PDFs into highly structured, machine-readable entity graphs.
The Execution: We executed a complete vector optimization of their core inventory (from local day-trips to multi-day European stays). By establishing distinct mathematical relationships between the “Città per te” entity, specific geographical nodes (e.g., Paestum, Amalfi, Amsterdam), and logistical data, we maximized their cosine similarity against high-intent prompts like “best organized group tours departing from Emilia-Romagna.”
The Output: The client’s full tour inventory — from local day-trips to multi-day European itineraries — is now structured for machine retrieval, with explicit semantic relationships between the “Città per te” entity, its geographical nodes and its logistical data (dates, pricing, departure points). This makes the catalogue retrievable and synthesisable by AI travel planners, where it was previously unreadable. Quantified Share of Model measurement against the pre-intervention state is scheduled.
ABN: 88 718 123 724
© 2026 All Rights Reserved
