The platforms that crawl your AI site the most cite you the most (except one)

Entry #5: The platforms that crawl your AI site the most cite you the most (except one)

Mar 3, 2026

Fresh data from citation tracking tests on March 2 shows 24 queries were run across four AI platforms. The tests measured how many times Genymotion was mentioned (brand mentions) and how many times sources linked to genymotion.com (citations). The results were compared to traffic measured in crawler logs on the AI site (rozz.genymotion.com). The article previously called this a “mirror site.” The author drops that term. The AI site is built for a different audience: AI agents. The AI site content is the same as a mobile site in content, but the audience and artefact are different. The article states it is not optimizing for citations, but for AI agents.

Key Findings

ChatGPT has an 83% citation rate, up from 14% before the AI site launched. This increase happens by the AI site providing structured content for AI agents.
Claude has a 21% citation rate. ClaudeBot crawl activity jumped 24x in one week (21 to 505 requests). The increase followed activation on Feb 28.
Perplexity has a 17% citation rate. PerplexityBot tripled week over week (14 to 42). The growth is early but on the same curve.
Gemini has a 4% citation rate despite recommending Genymotion in most queries. Gemini uses a different pipeline entirely.

The correlation (and where it breaks)

Here is what was measured on March 2, against crawl data from the same period.

| Platform | Crawling the AI site? | Weekly crawl volume | Citation rate | Brand mentioned | |---|---|---|---|---| | ChatGPT (GPT-5.2) | Heavy, sustained since January | 1,200+/week | 83% (20/24) | 96% (23/24) | | Claude | Just activated (Feb 28) | 505 this week, was 21 | 21% (5/24) | 33% (8/24) | | Perplexity (sonar-pro) | Light, growing | 42 this week, was 14 | 17% (4/24) | 25% (6/24) | | Gemini | Not crawling | 0 | 4% (1/24) | 38% (9/24) |

For ChatGPT, Claude, and Perplexity, more crawling corresponds to more citations. The AI site approach works for these platforms. Gemini behaves differently because it ignores the AI site entirely.

ChatGPT: 14% to 83%

Before building the AI site, genymotion.com appeared in roughly 14% of relevant AI queries. Eight weeks later, the rate rose to 83%. ChatGPT cites Genymotion in 20 of 24 use-case queries tested. Brand is mentioned in 23 of 24. The only thing that changed is the AI site with structured content for AI agents.

Recap: GPTBot’s initial mass crawl (547 requests on January 7) was followed by sustained ChatGPT-User traffic of 1,200+ requests per week. This week it reached 1,228 requests. The majority of requests hit Q&A pages (714), the homepage (307), and GEO content pages (150). The site moved from training to indexing to citing.

The pages cited are consistently on topics such as pricing plans (22 visits this week), macOS compatibility (13), free version availability (13), and Google Play Store setup (13). When ChatGPT recommends Genymotion, it links directly to genymotion.com pages. In most queries, Genymotion ranks at position #1. There are often multiple citation links per response. This reinforces legitimacy and makes ChatGPT’s answer a warm referral.

The remaining work is expanding content coverage into use cases where the site is weaker. CI/CD and app performance monitoring both show 13% citation rate, compared to 63% for app development and manual testing.

Claude: reading the map, not the pages

Two weeks ago, entry #4 noted waiting for ClaudeBot to return. ClaudeBot returned with 505 requests this week, up from 21, which is 24x. The important part is what ClaudeBot chose to read. GPTBot performs mass crawls: hundreds of pages per day, content first, structure optional. ChatGPT-User hits individual Q&A and GEO pages during live conversations—176 Q&A pages in the last four days. Both treat the AI site as a content pool to pull from. ClaudeBot reads the site’s organizational structure.

The topic sweep

On March 2, between 10:34 and 10:39 AM, ClaudeBot crawled 13 topic pages in sequence—one every 20–30 seconds. The topics include: android-versions, hardware-architectures, virtualization-technologies, arm-platform-and-gpu, billing-and-subscriptions, licensing-and-eulas, macos-security-toolkit, documentation-and-support, ci-cd-tools, network-security-toolkit, system-image-and-bios, software-installation-and-trials, root-access-and-tools.

ClaudeBot had previously done something similar on February 28: a session mixing topic pages with GEO content pages (enterprise features, cloud platforms, VirtualBox compatibility, Hyper-V troubleshooting), for eight content pages plus one topic page in two minutes.

ClaudeBot is the only bot that systematically crawls topic pages. In the same four-day window, ChatGPT-User hit two topic pages (both pulled into live conversations by user queries). GPTBot, ByteSpider, PerplexityBot, and Meta AI did not hit any topic pages.

What does “reading the map” actually mean?

Topic pages are the only pages on the AI site with CollectionPage schema. Each one lists every content page and Q&A in that topic, with titles, descriptions, and links. They function as the site’s table of contents: not answers to questions, but a map of what the site knows and how it is organized.

By crawling all 13 topic pages, ClaudeBot gains a complete picture of the site’s knowledge structure. ClaudeBot knows which topics the domain covers, how many pages address each topic, with which keywords, and how pages relate to each other.

What cannot be explained is why ClaudeBot read zero Q&A pages. During the same time, ChatGPT-User read 176 Q&A pages. Q&A pages are phrased as questions—the same format users use to query AI. If ClaudeBot has not indexed them, ClaudeBot lacks the content format that most directly matches user questions. If ClaudeBot returns for a Q&A crawl and ClaudeBot’s citation rate jumps, that would confirm this.

Three possible explanations

1) ClaudeBot is evaluating before committing. The pattern is constant monitoring (robots.txt and sitemap checks every 2–3 hours), then a content sample (Feb 28 session, 8 pages), then a structural assessment (Mar 2 topic sweep). This resembles a staged decision pipeline. The site’s scope and quality are assessed through its taxonomy before deeper crawls begin. If that is correct, a bulk content crawl should follow.

2) Anthropic is building a different kind of index. Claude’s retrieval may not follow ChatGPT’s “find page, index page, serve page” model. Anthropic could be building a knowledge-graph index where the unit of understanding is “what does this domain know about” rather than “what does this page say.” In that scenario, topic pages become the preferred starting point.

3) The Q&A gap explains the citation gap. Claude’s citation rate is 21%. ChatGPT’s is 83%. ChatGPT-User reads Q&A pages constantly. ClaudeBot has not indexed Q&A pages, which matches content formats users directly ask questions about. If ClaudeBot returns for a Q&A crawl and ClaudeBot’s citation rate increases, this would support the hypothesis.

The crawl itself

Outside content sessions, ClaudeBot runs a monitoring loop: robots.txt plus sitemap.xml every 2–3 hours, always from the same IP. One instance, checking in regularly, reading selectively when it decides to.

The citation test reflects this early stage. Claude cited Genymotion with a link in only 5 of 24 queries (21%), including a #1 position for “how can support agents replicate mobile app bugs on a virtual device” and #5 for manual testing. Brand mentioned in 8 of 24.

The pipeline difference

Three of four queries: Gemini provides answers from training data (no links). The user reads “get Genymotion” and must search for it themselves. One of four queries (pricing): Gemini triggers a live Google Search, grounds its response in web results, and cites through Google’s Vertex AI redirect URLs. That is the only query where genymotion.com receives a link.

What this means for GEO strategy: ChatGPT, Claude, and Perplexity have their own AI crawlers and retrieval indices. They fetch and cite the AI site content with direct links. Gemini does not have its own AI crawler. It either uses training data (no links) or triggers a Google Search (links go to Google results). The AI site is not present in Gemini’s pipeline at all because Gemini’s pipeline is Google-based.

For Gemini, the path to citations is not improved AI site content. It relies on Google Search ranking. GEO strategy requires two tracks: one for crawl-and-cite platforms (ChatGPT, Claude, Perplexity) and a separate one for Gemini, where the rules are different.

Where we are

Four platforms, four different stages:

ChatGPT: solved. 83% citation rate. Direct links to genymotion.com. The AI site approach works. Remaining work is expanding content coverage into weaker use cases.
Claude: structurally aware, content-light. 21% citation rate, growing. ClaudeBot is the only bot reading the site’s topic taxonomy. It may be building a structural model of domain expertise rather than a page-level retrieval index. Will it come back for a deep content crawl?
Perplexity: promising. 17% with light but growing crawl activity. The architecture is similar to the platforms where the AI site has proven it works. Deep indexing may follow the ChatGPT curve, but no deep indexing event has occurred yet.
Gemini: different game. Recommends Genymotion in most queries but does not link. Citations only happen when Gemini triggers a Google Search. The AI site does not reach Gemini’s pipeline. This might require a completely different approach.

One infrastructure handles three of four platforms. The fourth platform requires a different approach. Google SEO remains a viable channel for citations and discovery.

Data source

CloudFront access logs for rozz.genymotion.com, February 24 – March 3, 2026 (crawl data). Citation tests conducted March 2–3, 2026: 24 queries tested on ChatGPT, Claude, Perplexity, and Gemini. Bot classification based on User-Agent strings.

Author

Adrien Schmidt, CEO, ROZZ. A serial tech entrepreneur with 10+ years of experience building AI systems. Relevant projects include Aristotle (conversational AI analytics) and products for eBay and Cartier. The article notes prior ventures such as Squid Solutions and an AI products line for Cartier.