Which Pages of Corporate Websites Do Generative AI Cite? Optyino.ai Analyzes 35,510 AI Responses, Finding Homepage Citation Rate at 26.99%

Optyino.ai is a GEO/AIO/LLMO platform that visualizes and optimizes brand visibility in the age of AI search. Leveraging accumulated AI response logs, the company analyzed which types of pages (e.g., columns/articles, homepages, pricing, case studies) generative AI tends to cite from corporate websites.

This study analyzed 35,510 AI responses collected between August 11, 2025, and July 10, 2026, based on 40 prompts spanning diverse industries such as finance, insurance, accounting, education, real estate, B2B services, and e-commerce. Sample prompts included: 'Where is the best place to consult about tsutome NISA?', 'Which tutoring school is recommended for my elementary school daughter who isn’t particularly good at studying but I want her to try for middle school entrance exams?', and 'Which company is recommended for end-to-end apartment construction and management?'

Survey Overview

A total of 321,313 cited URLs were extracted from the 35,510 AI responses.

Excluding URLs from social media, e-commerce marketplaces, and comparison/review sites, URLs determined to be operated by the companies themselves were defined as 'official site URLs' in this report. There were 284,259 official site URLs, accounting for 88.47% of all cited URLs (this ratio is referred to as the 'official site URL rate').

Additionally, 32,721 responses (92.15% of 35,510) contained at least one official site URL (this ratio is referred to as the 'official site citation rate').

When categorized by page type, 'columns and articles' had the highest share among nine predefined categories, with a composition ratio of 35.95% and a response citation rate of 73.99%—significantly higher than other categories.

In contrast, the homepage had a composition ratio of 5.64% and a response citation rate of 26.99%. Pricing pages, which directly influence purchasing decisions, had a composition ratio of 2.51% and a response citation rate of 15.00%. Case study or results pages had even lower figures: 0.53% composition and 3.86% citation rate—highlighting a significant gap compared to columns and articles.

By AI model, the official site URL rate varied by 8.21 percentage points, from 91.95% for Gemini to 83.74% for ChatGPT. By industry, the official site citation rate was highest in real estate (95.47%) and relatively low in products/e-commerce (86.76%).

Background of the Survey

Whether a corporate official site is cited in a generative AI response is a key metric from the perspectives of GEO (Generative Engine Optimization), AIO (AI Optimization), and LLMO (Large Language Model Optimization). However, even when an official site is cited, the impact on user behavior—such as inquiries or purchases—differs significantly depending on whether the cited page is the homepage, a pricing page, or a column article.

Optyino.ai continuously collects and analyzes cited URLs from AI responses. This study focused specifically on 'official site URLs'—those determined to be on a company’s own domain—and classified them by URL structure (e.g., /price/, /case/, /faq/, /column/, /service/) to identify which types of pages within official sites are most frequently cited by generative AI.

Key Findings

36% of Official Site URLs Are 'Columns and Articles'—Generative AI Prefers 'Readability' Content

Among the nine predefined categories, 'columns and articles' had the highest composition ratio at 35.95%.

36.99% of official site URLs did not fall into any of the nine predefined page types and were categorized as 'Others'.

Since 'Others' is a residual category without specific functional roles, when comparing only the predefined categories, 'columns and articles' stand out with 35.95%, far exceeding the homepage at 5.64%—a more than sixfold difference. Other categories generally remained around 5% or lower.

This suggests that generative AI prefers citing in-depth column or article content over summary pages like homepages when explaining a company’s reality or specific information.

For companies enhancing their GEO strategy, consistently publishing thematic column articles may offer more citation opportunities than simply increasing content volume on the homepage.

Column/article response citation rate: 73.99%, homepage: 26.99%—pricing at 15%, case studies at just 3.9%

Among the nine predefined categories, 'columns and articles' had the highest response citation rate at 73.99%, significantly surpassing the homepage's 26.99%.

Overall URL Composition

Official Site URL Composition

Response Citation Rate

Columns/Articles

31.80%

35.95%

73.99%

Homepage

4.99%

5.64%

26.99%

Services/Features

4.07%

4.60%

25.49%

Articles/Detail Pages

3.74%

4.23%

24.86%

Locations/Regions

3.69%

4.18%

23.26%

The 'Others' category includes URLs not fitting predefined classifications. While its response citation rate is numerically highest at 82.89%, it cannot be directly compared as it lacks specific functional page groups.

Among predefined categories, the homepage's 26.99% is only about 36.5% of the column/article rate. Pricing pages, crucial for purchase decisions, reached only 15.00%, and case study pages just 3.86%.

This suggests that information companies most want to highlight—such as pricing and case results—may be inherently less likely to be cited by generative AI.

Rather than solely enriching pricing or case study pages, companies should consider creating separate explanatory column articles and designing pathways from these columns to pricing and case pages to increase citation opportunities.

Full data including FAQ/support, pricing, company information/trust, and case studies is available here:

https://optyino.ai/press/official-site-page-type-citation-rate