I-O-X Launches New 'LLMO Diagnosis' Service to Visualize Company Citations and Recommendations in AI Search

I-O-X Inc. (Headquarters: Shibuya, Tokyo; Representative Director: Tomonori Takibi), which has been operating SEO consulting and the search marketing media 'SEO Japan' for over 20 years, launched 'LLMO Diagnosis (Web Content Diagnosis)' on June 10, 2026. This service visualizes whether a company is cited or recommended in AI searches such as ChatGPT, Gemini, AI Mode, and AI Overviews.

## Google Begins Providing AI Search Exposure Data; LLMO Enters the 'Measure → Improve' Phase

On June 3, 2026, Google announced that it would add a 'Generative AI Performance Report' to Search Console to allow companies to check their exposure in AI searches. This report allows users to confirm the number of times (impressions) their pages have been displayed as citations or references in AI Overviews, AI Mode, and the generative AI frames in Discover, broken down by page, country, device, and date. The rollout began with select site operators in the UK and will gradually be expanded globally.

While currently focused on exposure (impressions) with limitations such as not being able to capture clicks or specific search queries, the significance of Google stepping into AI search visualization and data provision for site operators is immense. It signals that AI search is being positioned as an 'independent channel to be measured with dedicated metrics,' distinct from traditional organic search.

Until now, exposure in AI search had to rely on estimates from third-party tools, making it difficult to base decisions on a solid foundation. With official data now becoming available, LLMO is entering a stage where it can be tackled under the premise of a 'measure and improve' cycle (PDCA), just like search rankings and traffic.

In fact, when I-O-X analyzed 48 total responses from four major AIs (ChatGPT, Gemini, Google AI Mode, and AI Overviews) using 12 high-purchase-intent prompts related to considering an SEO firm, it was confirmed that AI recommendations are concentrated on a few companies, many companies never even appear as candidates, and that almost no responses cite the companies' own websites (owned media) as the source—with I-O-X itself being no exception.

Based on the know-how gained from tackling LLMO internally, 'LLMO Diagnosis' is a service that shows the priority of 'what should be addressed first.'

## What is 'LLMO Diagnosis (Web Content Diagnosis)'?

LLMO Diagnosis is a web content diagnostic service that examines both the inside of a company's own site (on-site) and external sources such as comparison sites and UGC platforms (off-site), clarifying the 'current position for being cited and recommended by AI' and the 'measures that should be addressed with priority.'

It prompts the context necessary for sales promotion and visualizes the web presence in the context that should be discussed.

## [Feature] A Proprietary Three-Axis Frame Back-Calculated from the Three Essential Steps of AI Search

The process by which AI generates responses can be broken down into three stages: ① Decomposition → ② Retrieval → ③ Synthesis.

LLMO Diagnosis stands on the premise that if the materials needed by AI at each stage are back-calculated and prepared, the probability of the company being cited and recommended can be increased. It evaluates and designs based on three axes corresponding to this immutable structure:

- Comparison Design (Decomposition): Does the company emerge as a candidate in arbitrary sub-queries where AI decomposes questions?
- Referencability (Retrieval): Is it referenced as a verifiable information source on a passage-by-passage basis?
- Citation Ease (Synthesis): Is it written in a way that is easy to extract and identifies 'whose, what, and which product' it is?

The feature of this service is that it is designed based on the immutable structure of AI search, without spending hours on superficial measures with short shelf lives.

*In its design, the service references Google's 'Thematic Search Patent (US12158907B1)' and Google DeepMind's 'AlphaEvolve (arXiv:2506.13131, 2025).'

## LLMO Diagnosis Service Details

Answering 'In the end, what should I do?' with a rationale:

- Know Your Current Position: Measure your company's mention/citation status in AI search and share with competitors to show exactly where you stand in numbers.
- Clarify Gaps with Three-Axis Diagnosis: Analyze where gaps exist across the three axes of comparison design, referencability, and citation ease.
- Propose Priority Measures: Present priority-based suggestions on 'what to address first' to achieve results with limited resources.