Entry #3 · Feb 17, 2026

3x Week-Over-Week: What Sustained ChatGPT Citation Growth Looks Like

ChatGPT citations on rozz.genymotion.com hit 1,077 this week—3x the 345 citations we reported last week. Last week we shared data showing 16x citation growth in 7 days. This week it tripled again. 1,077 ChatGPT-User requests between Feb 10–17, up from 345 the week before. The flywheel is working. Every week, ROZZ crawls Genymotion's main site for updates, processes 500+ new questions from the chatbot, and publishes fresh Q&A pages to the mirror site. Six weeks ago, ROZZ recorded 42 citations in an entire month. ROZZ now recorded 1,077 in a single week.

Key Findings

1,077 ChatGPT-User citations in 7 days (Feb 10–17)—3x the previous week's 345
Six-week trajectory: 42 → 345 → 1,077 citations
Q&A pages remain the citation driver: 66% of all ChatGPT requests
PerplexityBot activity up 5x: 33 requests vs 7 the previous week
Total LLM bot requests: 2,606 in 7 days
GPTBot still actively training: 625 requests this week

The Growth Trajectory

| Period | ChatGPT Citations | Change | |---|---:|---| | Jan 3 – Feb 2 (30 days) | 42 | Baseline | | Feb 2–9 (7 days) | 345 | +721% | | Feb 10–17 (7 days) | 1,077 | +212% |

From 42 to 1,077 in six weeks. That is 25x growth—with no additional work after the initial mirror site deployment.

Daily Breakdown (Feb 10–17)

| Date | Citations | |---|---:| | Feb 10 | 107 | | Feb 11 | 131 | | Feb 12 | 109 | | Feb 13 | 93 | | Feb 14 | 141 | | Feb 15 | 252 | | Feb 16 | 196 | | Feb 17 | 48* |

*Feb 17 is partial day (data through ~12:30 UTC)

The pattern shows sustained high volume with a spike on Feb 15. Feb 15 had 252 citations in one day. Feb 15 exceeded the entire previous week's total from just a month ago.

What's Getting Cited

Q&A pages continue to dominate. Of the 1,077 ChatGPT-User requests:

| Content Type | Citations | Share | |---|---:|---:| | Q&A Pages | ~710 | 66% | | Homepage | ~250 | 23% | | GEO Pages | ~100 | 9% | | Other | ~17 | 2% |

The same pattern holds: structured Q&A content with Schema.org QAPage markup gets cited far more than traditional content pages.

The Full Bot Picture

ChatGPT-User citations are the outcome. The full ecosystem of AI activity this week is:

| Bot | Requests | Purpose | |---|---:|---| | ChatGPT-User | 1,077 | Live citations in user conversations | | CCBot | 710 | Common Crawl (training data) | | OpenAI GPTBot | 625 | GPT training data | | Meta AI | 70 | Llama training data | | OpenAI SearchBot | 67 | Building retrieval indexes | | PerplexityBot | 33 | Perplexity indexing | | ByteSpider | 18 | ByteDance/TikTok training | | ClaudeBot | 6 | Claude training data | | Total | 2,606 | — |

Notable this week: PerplexityBot activity jumped from 7 to 33 requests. Still small, but a 5x increase suggests Perplexity is starting to index the mirror site more seriously.

Why It Keeps Growing

The mirror site infrastructure creates a flywheel:

1. Fresh content every week ROZZ crawls Genymotion's main site weekly for updates. ROZZ processes 500+ chatbot questions. New Q&A pages get published automatically. The content is tuned to exactly what users are asking right now.

2. More content gets indexed over time GPTBot made 625 requests this week. Every crawl adds more content to training data and retrieval indexes. The citation surface area expands.

3. Successful citations reinforce the source When ChatGPT retrieves content that answers a user's question well, that signals quality to the retrieval system. Good answers beget more retrieval.

4. Q&A pages match query patterns The Q&A pages on the mirror site are phrased as questions—exactly how users ask ChatGPT. "What are Genymotion's pricing options?" matches "what are genymotion pricing options" naturally.

5. Schema.org markup makes extraction trivial Every Q&A page has QAPage structured data. The question and answer are explicitly marked. AI systems don't have to parse marketing copy—they get clean, structured content.

Six Weeks of Data

| Week | ChatGPT Citations | Cumulative | |---|---:|---:| | Jan 3 – Feb 2 | 42 | 42 | | Feb 2–9 | 345 | 387 | | Feb 10–17 | 1,077 | 1,464 |

1,464 total citation events in six weeks. Real users. Real questions. Real answers sourced from the mirror site. And the trajectory is accelerating, not flattening.

What ROZZ Built

The mirror site at rozz.genymotion.com includes:

450 GEO pages: AI-optimized versions of help center articles and documentation
177 Q&A pages: Generated from real user questions via the ROZZ chatbot
Schema.org markup on every page: QAPage for Q&As, WebPage for content
llms.txt discovery files: Index and full-content versions
JSON APIs: Programmatic access for AI systems

This infrastructure was built once. The citations keep compounding.

Genymotion is one client. ROZZ builds this for every domain.

What This Means for GEO

The timeline is fast. Six weeks from deployment to 1,000+ weekly citations. GEO shows results in weeks, not the months typical of traditional SEO.

Growth compounds. You don't need to keep adding content to see citation growth. The same infrastructure generates more citations as AI systems index deeper and retrieval improves.

Q&A content wins. 66% of citations hit Q&A pages. Structured question-answer content with Schema.org markup outperforms traditional pages by a wide margin.

Multiple AI ecosystems. It is not just ChatGPT. GPTBot, ClaudeBot, Meta AI, PerplexityBot, CCBot—the mirror site feeds them all simultaneously. One infrastructure, multiple AI platforms.

Get This for Your Site

ROZZ builds this infrastructure automatically. Mirror site. Q&A pages from your chatbot. Schema.org markup on every page. llms.txt discovery files. The complete AI publishing layer.

$997/month | 1,077 citations in a week

→ Book a call | → See how it works | → rozz@rozz.site

Data source: CloudFront access logs for rozz.genymotion.com, February 10–17, 2026. Bot classification based on User-Agent strings.

Serial tech entrepreneur with 10+ years experience building AI systems including Aristotle (conversational AI analytics) and products for eBay and Cartier.