Entry #4 · Feb 24, 2026

Bing just found Genymotion: 1,556 BingBot requests and what it means

This week, BingBot made 1,556 requests to rozz.genymotion.com. This week included more requests than any other bot. ChatGPT-User was included among those other bots. We were watching ChatGPT. The citation numbers had been growing fast. The citation count was 42 in January. The citation count was 345 the first week of February. The citation count was 1,077 the week after. We expected this week to be another ChatGPT chapter. Instead, BingBot showed up at a scale we hadn’t seen before. 1,556 requests in 7 days was more than GPTBot (129). 1,556 requests in 7 days was more than ByteSpider (1,225). 1,556 requests in 7 days was more than ChatGPT-User citations (1,329). What started as a ChatGPT story is now happening across six platforms.

Key Findings

### BingBot
BingBot requests were 1,556 (Feb 17–24).
BingBot was the largest single bot category this week.

### ChatGPT-User citations
ChatGPT-User citations were 1,329.
ChatGPT-User citations were up from 1,077 the prior week.
The change was (+23%).

### Total LLM bot requests
Total LLM bot requests were 3,188.

### Six platforms now crawling the mirror site
OpenAI
Microsoft
Anthropic
Meta
ByteDance
Perplexity

The Data

All bots (Feb 17–24, 2026)

| Bot | Requests | Purpose | Platform | |---|---|---|---| | BingBot | 1,556 | Search index | Microsoft | | ChatGPT-User | 1,329 | Live citations | OpenAI | | ByteSpider | 1,225 | Training | ByteDance | | CCBot | 382 | Training | Common Crawl | | OpenAI GPTBot | 129 | Training | OpenAI | | OpenAI SearchBot | 75 | Search index | OpenAI | | ClaudeBot | 21 | Training | Anthropic | | PerplexityBot | 14 | Search index | Perplexity | | Meta AI | 13 | Training | Meta | | Total | 3,188 | — | — |

ChatGPT-User citation growth

| Period | Citations | |---|---| | Jan 3 – Feb 2 (30 days) | 42 | | Feb 2–9 (7 days) | 345 | | Feb 10–17 (7 days) | 1,077 | | Feb 17–24 (7 days) | 1,329 |

What BingBot is and why its crawl matters

BingBot feeds three Microsoft AI products. Copilot is the AI assistant in Windows 11 and Microsoft 365. Bing AI is the search experience in Edge. Azure OpenAI is used by enterprise customers to run hosted GPT models against a live web index.

When BingBot indexes a page, that content enters the retrieval layer for all three products. 1,556 requests is the same training-to-citation pipeline we’ve been watching with OpenAI. The pipeline now starts over at Microsoft scale. The pipeline is training → indexing → live citation.

The requests were spread across all 7 days. The requests covered both Q&A pages and GEO content pages. We don’t yet have a Copilot-User equivalent. The Copilot-User equivalent would confirm live citations the way ChatGPT-User does for OpenAI. Based on what we saw with OpenAI, we’d expect Copilot citations to start showing up around mid-March.

What’s being cited

The questions getting cited are consistent week over week. The cited questions include system requirements. The cited questions include macOS compatibility. The cited questions include pricing. The cited questions include Play Store setup. The cited questions include how to download the free version. These cited questions are the questions someone asks when they’re deciding whether to use the product. These cited questions are not after someone has already signed up.

| Top cited Q&A pages | Bots | |---|---| | What are the system requirements? | ChatGPT-User, ByteSpider | | How do I download Genymotion Desktop and what are the system requirements? | ChatGPT-User, ByteSpider | | Does the emulator work with the latest macOS? | ChatGPT-User, ByteSpider | | What pricing plans are available? | ChatGPT-User, ByteSpider | | Is Genymotion free to use? | ChatGPT-User | | How can I enable Google Play Store on Genymotion? | ChatGPT-User |

The two that haven’t followed through: ClaudeBot and PerplexityBot

Not everything is growing. ClaudeBot and PerplexityBot are genuinely puzzling.

ClaudeBot had a real crawl on February 3. ClaudeBot went through roughly 60 pages and Q&As in one session. ClaudeBot’s burst pattern was similar to what we saw with GPTBot in January. ClaudeBot stopped after that session. Since then, robots.txt and the homepage were the only pages visited. Every few days, nothing else was found. There was no follow-up content crawl. There was no Q&A deep dive. Nothing resembling the secondary and tertiary waves we saw from GPTBot on January 18–19 and January 25–26 appeared.

Three weeks of waiting occurred. ClaudeBot had no return.

PerplexityBot is even lighter. PerplexityBot has been visiting since early February. PerplexityBot visited only 1–4 pages per session. PerplexityBot always visited the same small set. That set included a rooting guide. That set included the Android 16 beta post. That set included the SaaS product page. That set included the system requirements page.

PerplexityBot hasn’t tried the Q&A index. PerplexityBot hasn’t touched llms.txt. PerplexityBot hasn’t done anything that looks like systematic indexing.

Total PerplexityBot requests were 14 this week. Total PerplexityBot requests were 30 over 60 days.

> What we’re testing next: The content on the mirror site is the same content that GPTBot and BingBot found worthwhile. > Q&A pages are Schema.org-marked. > The llms.txt file is there. > The sitemap is clean. > ClaudeBot read the page. > ClaudeBot found 60 pieces of content worth crawling. > ClaudeBot then apparently decided it had seen enough. > We plan to make llms-full.txt more prominent at the root of the website and report back.

What to watch for next

Three weeks after GPTBot’s mass crawl in January, ChatGPT-User citations started appearing. BingBot just ran a similar mass crawl.

If the same timeline holds, Copilot citations should start showing up around March 10–17.

That would mean two independent citation pipelines running from the same content. OpenAI and Microsoft would each run a pipeline. Both pipelines would be built on the same mirror site infrastructure. One architecture would serve both pipelines. Two platforms would cite that same content. We’ll report back when we see it.

Get This for Your Site

ROZZ builds this infrastructure automatically. A mirror site is used. Q&A pages are generated from your chatbot questions. Schema.org markup is applied on every page. llms.txt discovery files are included.

$997/month | One architecture, six platforms indexing

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

Data source: CloudFront access logs for rozz.genymotion.com, February 17–24, 2026. Bot classification is 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. Previously founded Squid Solutions and built AI products like Aristotle, the conversational big data analytics chatbot, and an AR jewelry try-on device for Cartier.

February 24, 2026 | Data period: Feb 17–24, 2026