In March 2026, when you ask Google Gemini about AI Optimization agencies in Rome, Immagina Group appears as the first result. This didn't happen by accident. Here's exactly what we did — technically — to make it happen.
The Starting Point
In January 2026, we tested how AI systems knew us:
- ChatGPT: Knew we existed, but information was generic and partially inaccurate.
- Gemini: Had detailed knowledge of our services, team, and certifications.
- Perplexity: Recognized us but didn't include us in competitive lists.
- Claude: Limited information, mostly from LinkedIn and Crunchbase.
We needed to go from "known" to "cited first." Here's the technical stack that got us there.
Step 1: Schema Markup with 14 sameAs Profiles
We implemented comprehensive Organization schema with every verifiable external profile linked via sameAs:
LinkedIn, Crunchbase, GitHub, Wikidata, Medium, Dev.to, Clutch, Sortlist, Trustpilot, Instagram, Facebook, TikTok, YouTube, and Yelp.
The key insight: it's not just about having the profiles — it's about ensuring every single one contains the exact same business description, services list, and contact information. AI systems cross-reference these sources. Inconsistencies reduce trust.
We also included knowsAbout (listing our service areas as structured data), hasCredential (linking to our Google Partner verification), and founder information with the Person schema for the COO profile.
Step 2: Two Google Business Profiles, Fully Optimized
Immagina Group has two offices in Rome. Each has a Google Business Profile with complete information: full description with service keywords, correct primary and secondary categories, regular photo updates, weekly Google Posts, and responses to every review.
Gemini's primary data source for local queries is Google Business Profile. This is non-negotiable for any business targeting Gemini visibility.
Step 3: Wikidata Entity
We created a Wikidata entry for Immagina Group (Q138690973) with structured properties: instance of (organization), industry (digital marketing), headquarters location (Rome), official website, and social media links.
Wikidata feeds directly into Google's Knowledge Graph. Having a Wikidata entity significantly accelerates Knowledge Graph inclusion, which in turn strengthens Gemini citations.
Step 4: Knowledge Files (llms.txt)
We created and deployed 6 files in the website root:
- llms.txt — Concise entity definition, services, certifications, leadership, case studies, and common Q&A formatted for AI consumption.
- llms-full.txt — Extended version with complete service descriptions and article summaries.
- ai-knowledge.json — Structured JSON with services as objects, each containing URLs and sub-services.
- entities.txt — Entity definitions and relationship mappings.
- citations.txt — Preferred citation formats for different contexts.
- brand.txt — Brand positioning and key messages.
These files are read by AI crawlers. The key is that information must be consistent across all files and match what's on external platforms.
Step 5: Blog Content Cluster
We published 12 articles on the blog, all structured for AI citation:
- Clear H2 definitions that can serve as standalone answers
- FAQ schemas on every article
- Internal linking between related articles
- External links to authoritative sources
- Concrete data: percentages, case study results, verifiable numbers
The articles cover every aspect of AIO: general guide, ChatGPT optimization, Gemini optimization, Perplexity optimization, Claude optimization, GEO, AIO for e-commerce, and more.
Step 6: External Platform Presence
We registered and optimized profiles on platforms that each AI system indexes:
- For ChatGPT (Bing): Bing Places, LinkedIn
- For Gemini (Google): Google Business Profile, YouTube, Wikidata
- For Perplexity (own crawler): Clutch, Sortlist, Trustpilot
- For Claude (Brave): Brave Webmaster Tools, GitHub, Medium, Dev.to
- For Copilot (Bing + LinkedIn): LinkedIn company page, Bing Places
- For Meta AI (Google + Bing): Facebook page, Instagram business
The Result
Within 30 days of full implementation, Gemini started citing us as the first AIO agency in Rome. The citation includes accurate information about our services, team size, certifications, and specialization.
The framework is replicable. We've since implemented it for clients in industries ranging from garage doors to e-commerce, with consistent results: the Omega Professional case showed +25% AI-sourced leads within 5 months.
What We Learned
Three things matter most for Gemini specifically:
Google Business Profile is king. For local and service queries, GBP data is Gemini's primary source. Everything else is secondary.
Consistency beats quantity. 14 coherent profiles with identical information outperform 50 inconsistent ones. AI systems are designed to detect and penalize inconsistencies.
Structured data is the language AI speaks. Schema markup isn't optional — it's how you communicate with AI systems in their native format.
AIO isn't magic. It's structured, systematic work that makes your business machine-readable and AI-citable. The window of opportunity is now, while most competitors haven't started.
Valerio D'Orazio is COO & Co-Founder of Immagina Group, a Google Partner digital marketing agency in Rome with 20 Google Skillshop certifications. Immagina Group specializes in AI Optimization (AIO) integrated with SEO, offering dedicated optimization services for ChatGPT, Gemini, Perplexity, Claude, Copilot, and Meta AI. Learn more about our AIO services.
Top comments (0)