Even with Top SEO Rankings, AI Won't Cite You. IDEATECH Launches "IDEA LLMO," a B2B-Focused LLMO Support Service to Gain AI Recommendations Starting with Third-Party Mentions

IDEATECH (Headquarters: Minato-ku, Tokyo; President and CEO: Tomoo Ishikawa), a company specializing in B2B strategic PR, has launched "IDEA LLMO," a B2B-focused LLMO (Large Language Model Optimization) support service. This service aims to create a state where companies are recommended and cited by conversational generative AI (such as ChatGPT, Gemini, Google AI Overview, etc.).

Based on over 15 years of strategic PR experience and planning capabilities in over 2,500 B2B-focused research PR projects, we design the acquisition of "third-party mentions (citations)" that AI preferentially references. Furthermore, drawing on our strategic PR expertise, we conduct information design and content design, including determining which categories (prompts) to target and how to create information that AI can trust (discovering strengths).

■ Background of Service Provision

B2B purchasing behavior is shifting from search engine queries to consultations with conversational generative AI. A survey by Forrester indicates that by 2025, 94% of B2B buyers will use generative AI in their purchasing process, and responses citing generative AI and conversational search as "most valuable information sources" were twice as high as any other source, including vendor sites and sales representatives. A G2 survey also found that 51% of B2B software buyers start their product research with AI chatbots, and 69% select vendors different from their initial plan based on AI suggestions. Buyers are beginning to "describe their situation in text to have AI narrow down options," a highly granular need that traditional search keywords cannot address.

This shift cannot be addressed by measures extending from conventional SEO. An analysis of approximately 40,000 queries by Moz revealed that 88% of links cited by Google AI Mode were not in the top 10 search results for the same keywords. This phenomenon, where high search rankings do not lead to AI citations, is occurring. Another Ahrefs study showed that only 38% of AI Overview citations matched the top 10 search results, a halving from 76% in July 2025. The correlation between search rankings and AI citations is breaking down.

Crucially, conversational generative AI preferentially references information mentioned by third parties over content published by the company itself, considering it more reliable. A Muck Rack survey found that approximately 89% of links cited by ChatGPT and Gemini were third-party information, such as news articles, specialized media, and reviews. Furthermore, an Ahrefs analysis of 75,000 brands revealed that the factor most strongly correlated with brand exposure in AI answers was "brand mentions on third-party websites (citations)" (correlation coefficient 0.664), approximately three times stronger than backlinks (0.218), which are a core asset of conventional SEO. No matter how well a company optimizes its own website, if it lacks mentions from third-party media, it will be difficult for that information to be reflected in AI responses.

Companies recommended by AI possess the proof of trust from "AI, a third party." Companies named by AI begin negotiations as "already trusted candidates," while those not named are not even considered for comparison.

▼ IDEA LLMO Service Site: https://ideatech.jp/service/idea-llmo/

■ 86.6% of LLMO Personnel Experience Challenges; About 60% of Those Who Outsource Find "Insufficient Factual Information"

An independent survey conducted by IDEATECH yielded the following results:

Reference: Survey on the Selection of LLMO Support Companies in B2B Business

https://prtimes.jp/main/html/rd/p/000000576.000045863.html

86.6% of personnel working on LLMO measures perceive challenges and hurdles.

Approximately 60% (58.3%) of those with outsourcing experience responded that "the quality and quantity of factual information were insufficient." "Shallow understanding of B2B" and "lack of coordination between production and dissemination" also exceeded 50% each, making a system capable of consistently handling everything from initial research design to PR distribution a selection criterion.

The top priority in selection was "strategic proposal capabilities for utilizing third-party data" (39.5%), followed by "understanding of B2B business practices" (39.2%). Including "consistent support" (34.7%) as the third priority, companies capable of handling research, content, and PR distribution without fragmentation are in demand.

【Survey Overview】

Survey Name: Survey on the Selection of LLMO Support Companies in B2B Business

Survey Method: Internet survey planned by "Lisapy®︎," a research marketing service provided by IDEATECH

Survey Period: March 9-10, 2026

Valid Responses: 329 marketing managers and representatives working in companies with B2B operations who responded that they are involved in measures for conversational generative AI.

■ How to Choose an LLMO Support Company: 6 Judgment Criteria

Based on the survey results above, we organize the judgment criteria for selecting an LLMO support company. For B2B companies, the important judgment criteria are: "LLMO Technical Understanding," "Research Design Capability x Content Production Capability," "Ability to Acquire Mentions in Third-Party Media," "B2B/Industry Understanding," "Monitoring System," and "In-house Practice."

Judgment Criteria

Points to Check

Cautionary Notes

① LLMO Technical Understanding

Can they explain the mechanism by which AI learns and cites information?

There are cases where "We support LLMO" is just a slogan with no substance.

② Research Design x Content Production

Can they consistently handle everything from planning initial research to releases, columns, and white papers?

Relying on a research company may lack PR perspective; relying on a PR company may result in weak research design.

③ PR & Third-Party Media Acquisition

Do they have a track record of publishing research data in PR TIMES or industry media?

Simply publishing on their own website will not increase third-party citations.

④ B2B/Industry Understanding

Do they have support experience in the B2B/SaaS domain, and can they understand specialized terminology?

Companies focused on B2C results often cannot handle the complexity of B2B purchasing processes.

⑤ Monitoring System

Do they have a system for continuously measuring AI recommendation status, LLM traffic, and CV?

Ensure reports are not just "number of mentions" and can be used for strategy improvement.

⑥ In-house Practice

Does the support company itself implement and demonstrate LLMO measures in-house?

Services not tested internally have uncertain reproducibility.

■ Features of "IDEA LLMO" | Building Proof of Trust from Third Parties Through PR Strategy

"IDEA LLMO" starts not just with content creation, but with accumulating proof of trust from third parties. It consists of a three-step process to achieve AI recommendations: focusing on publicity, designing industry selection contexts, and continuous monitoring.

STEP 01 | Publicity Strategy: Strategically Acquire Mentions from Third-Party Media

We design research plans that media outlets find compelling and strategically accumulate third-party mentions that AI references through web media placements.

Measures: Strategic distribution of research releases / Industry media