We changed one page. PerplexityBot went from 42 requests to 511.

Source:

Published:

Keywords: PerplexityBot, OpenAI SearchBot, ChatGPT-User, ClaudeBot, GEO Pages, Q&A Pages, topic taxonomy

Document ID: https://rozz.site/insights/perplexity-wakes-up.html#split1

We changed one page. PerplexityBot went from 42 requests to 511.

Entry #7 · Mar 17, 2026

PerplexityBot made 511 requests to Genymotion’s AI site in the week of March 10–17, up from 42 the week before. That change was a 12x increase. PerplexityBot crawled 172 Q&A pages. PerplexityBot crawled 256 GEO pages. PerplexityBot swept the entire topic taxonomy twice. Five days earlier, Genymotion redesigned the site’s index page. Genymotion cannot prove a connection between the redesign and the increase. PerplexityBot had been stuck on the same four pages for six weeks. PerplexityBot then stopped visiting the same four pages.

For months, PerplexityBot was the bot that Genymotion could not figure out. PerplexityBot showed up in early February. PerplexityBot visited a rooting guide. PerplexityBot visited the Android 16 beta page. PerplexityBot visited the SaaS product page. PerplexityBot visited the system requirements page. PerplexityBot then came back and visited the same pages again. PerplexityBot then came back again and visited the same pages. Total across six weeks, PerplexityBot never made more than 42 requests in a given week. PerplexityBot made no Q&A deep dives. PerplexityBot made no topic pages visits. PerplexityBot did not do systematic indexing.

On March 9, Genymotion redesigned the AI site’s index page. On March 10, PerplexityBot ran its first real content crawl.

Key findings

The data

All LLM bots (Mar 10–17, 2026)

| Bot | Requests | Category | Change vs. prior week | | --- | --- | --- | --- | | ChatGPT-User | 1,176 | Citation | Stable (~1,200) | | PerplexityBot | 511 | index | 12x (was 42) | | OpenAI SearchBot | 328 | index | 4x (was ~75) | | OpenAI GPTBot | 171 | Training | Moderate growth | | ClaudeBot | 123 | Training | Down from 505 | | CCBot | 100 | Training | Stable | | ByteSpider | 69 | Training | Down | | Meta AI | 52 | Training | Stable |

PerplexityBot content breakdown

| Content type | Requests | % | | --- | --- | --- | | GEO Pages | 256 | 50.1% | | Q&A Pages | 172 | 33.7% | | Topic Pages + Other | 61 | 11.9% | | Robots.txt | 12 | 2.3% | | Homepage | 4 | 0.8% |

84% of PerplexityBot’s requests hit content pages. PerplexityBot skipped the homepage almost entirely. PerplexityBot went straight to the content.

What we changed

On March 9, Genymotion redesigned the AI site’s index page at rozz.genymotion.com.

The old index was an infrastructure page. The old index listed API endpoints. The old index listed content counts. The old index listed navigation links. The old index told a human reader “here’s 177 Q&A pages and 450 GEO pages, here’s how to browse them.” AI crawlers did not use JSON APIs. AI crawlers GET HTML pages. When a bot landed on the old index, the bot got a page that described the site’s architecture. The old index did not give any signal about what Genymotion was. The old index did not give any signal about which content to read first.

Genymotion had already identified this as a problem in Entry #6. Genymotion found that 28% of ChatGPT-User sessions hit the index and stopped. The index created dead ends. The bot arrived. The bot found no useful signal. The bot left.

The new index opens with a product description. The new index describes what Genymotion is. The new index describes what Genymotion does. The new index describes who uses Genymotion. After the product description, the new index provides a topic directory with inline summaries for each topic. The new index provides context about pricing tiers, free personal use, SaaS vs Desktop costs, and plan differences. The topic directory provides enough signal for a crawler to answer a query from the index itself. The topic directory provides enough signal for a crawler to know exactly which page to fetch next.

What PerplexityBot did after the change

Day 1: March 10, content burst

The day after the index revamp, PerplexityBot ran a content crawl at 21:03 UTC. PerplexityBot crawled 50+ pages in under 20 seconds. PerplexityBot crawled GEO pages. PerplexityBot crawled Q&A pages. PerplexityBot crawled ARM transition content. PerplexityBot crawled Android 16 pages. PerplexityBot crawled pricing. PerplexityBot crawled installation. PerplexityBot crawled virtualization. PerplexityBot crawled case studies. PerplexityBot crawled release notes. PerplexityBot came from 8 different IP addresses running in parallel.

This was the first time PerplexityBot had ever crawled more than 4 pages in a single session on the AI site.

Day 5: March 14, Q&A deep dive

PerplexityBot returned at 21:02 UTC. PerplexityBot crawled 40+ Q&A pages in a single burst. PerplexityBot fetched one page every 1–2 seconds.

The questions PerplexityBot grabbed included pricing. The pricing questions included how much it costs. The pricing questions included pricing limits. The pricing questions included a free trial. The pricing questions included discounts. The pricing questions included monthly cost.

The questions PerplexityBot grabbed included installation. The installation questions included Windows steps. The installation questions included Linux steps. The installation questions included macOS. The installation questions included Ubuntu. The installation questions included general setup.

The questions PerplexityBot grabbed included compatibility. The compatibility questions included MacBook Air 8GB RAM. The compatibility questions included ARM support. The compatibility questions included VirtualBox dependency. The compatibility questions included Windows 10 Intel.

The questions PerplexityBot grabbed included product questions. The product questions included 50 concurrent devices. The product questions included license keys. The product questions included online vs download.

This is the content format that drives 75% of ChatGPT citations. PerplexityBot had never touched a Q&A page before this week.

Day 6: March 15, topic sweeps + GEO content

PerplexityBot used two sessions. PerplexityBot did a sweep at 21:21. PerplexityBot swept 14 topic pages in rapid succession. The 14 topic pages were all within seconds of each other.

These were CollectionPage-marked pages that list every piece of content in a topic. CollectionPage-marked topics included android-development-toolkit. CollectionPage-marked topics included ci-cd-toolchain. CollectionPage-marked topics included cloud-platform-ecosystem. CollectionPage-marked topics included hypervisor-platforms. CollectionPage-marked topics included licensing-eula. CollectionPage-marked topics included saas-paas-tools. The list also included more topics.

PerplexityBot came back for more GEO pages at 22:59. PerplexityBot crawled ARM webinar recordings. PerplexityBot crawled CircleCI integration. PerplexityBot crawled mobile testing. PerplexityBot crawled VPN setup. PerplexityBot crawled proxy configuration. PerplexityBot crawled biometric authentication. PerplexityBot crawled Bitrise CI. PerplexityBot crawled Applitools visual testing.

Day 8: March 17, second topic sweep

PerplexityBot did another sweep of topic pages at 07:13. PerplexityBot crawled genymotion-virtual-devices. PerplexityBot crawled gpu-arm-hardware. PerplexityBot crawled technical-support-troubleshooting. PerplexityBot crawled android-testing-stack. PerplexityBot crawled network-system-security. PerplexityBot crawled marketplace-cloud-providers.

PerplexityBot reads the map and the pages

PerplexityBot is the second bot Genymotion has seen systematically crawl topic pages. ClaudeBot did this first on March 2. ClaudeBot swept 13 topic pages in five minutes.

PerplexityBot performed the topic sweep and the content deep dive. PerplexityBot read the map and then read the chapters. ClaudeBot read the map and stopped.

Here is how the bots compare on content type distribution this week.

| Bot | GEO Pages | Q&A Pages | Topic/Other | Homepage | | --- | --- | --- | --- | --- | | PerplexityBot | 50% | 34% | 12% | 1% | | OpenAI SearchBot | 45% | 24% | 16% | 3% | | ChatGPT-User | 9% | 34% | 20% | 37% | | ClaudeBot | 1% | > Warning: 0% | 64% | 5% |

PerplexityBot and OpenAI SearchBot share a similar profile. PerplexityBot and OpenAI SearchBot have content-heavy crawling. PerplexityBot and OpenAI SearchBot have minimal discovery overhead.

Both PerplexityBot and OpenAI SearchBot are index bots. Both PerplexityBot and OpenAI SearchBot skip the homepage. Both PerplexityBot and OpenAI SearchBot go straight to content pages.

ChatGPT-User uses the homepage as a discovery hub. ChatGPT-User sends 37% of requests to the homepage. ChatGPT-User then drills into Q&A pages. ChatGPT-User sends 34% of requests to Q&A pages.

This makes sense. ChatGPT-User operates during live conversations. ChatGPT-User needs a starting point.

ClaudeBot is in a category of its own. ClaudeBot has 64% of requests to robots.txt, sitemaps, and topic pages. ClaudeBot has 0% Q&A pages. ClaudeBot has 1% GEO content. ClaudeBot is still in discovery mode.

> Warning: 64% > Warning: 0% > Warning: 1%

The search index layer is catching up

PerplexityBot (511) and OpenAI SearchBot (328) together made 839 requests this week. That volume was 71% of ChatGPT-User’s citation volume (1,176).

For context, search index bots were negligible two months ago. SearchBot was at 66 requests in January. PerplexityBot was at 14 requests. Combined search index bots were 80 requests. Combined index traffic was 839 requests. The growth was a 10x increase.

| Week | Search index bots | ChatGPT-User citations | Ratio | | --- | --- | --- | --- | | Jan 3–Feb 2 | 80 | 42 | 1.9x | | Feb 17–24 | 89 | 1,329 | 0.07x | | Mar 10–17 | 839 | 1,176 | 0.71x |

In January, search index activity exceeded citation activity. ChatGPT citations then exploded. Search index bots then fell behind. Now the search index layer is catching back up.

> If the ChatGPT pattern holds (deep indexing followed by citations approximately three weeks later), Perplexity citations should start appearing around early April.

ChatGPT: steady state

ChatGPT-User made 1,176 requests this week. The trajectory appears below.

| Period | ChatGPT-User | Change | | --- | --- | --- | | January (30 days) | 42 | Baseline | | Feb 2–9 | 345 | 8x | | Feb 10–17 | 1,077 | 3x | | Feb 17–24 | 1,329 | +23% | | Mar 3–10 | ~1,200 | Stable | | Mar 10–17 | 1,176 | Stable |

The exponential phase is over. ChatGPT citations have plateaued around 1,200 requests per week. This is roughly 44 content sessions per day. Each content session averages 3.1 pages. 40% of sessions involve multiple turns. Users ask follow-up questions in the same conversation.

This week’s deepest session: 22 pages, 8 turns, 38 minutes. A user in the US-Central region started with macOS system requirements. The user drilled into Android versions. The user explored device templates. The user checked plugin setup. The user came back multiple times.

This is not a lookup. This is a full product evaluation happening inside ChatGPT. The AI site is the knowledge source.

The growth frontier has moved. ChatGPT is now a steady-state citation channel. The next growth story is Perplexity.

> The growth frontier has moved. ChatGPT is now a steady-state citation channel. The next growth story is Perplexity.

Claude: still waiting

In Entry #5, Genymotion wrote about ClaudeBot’s distinctive behavior. ClaudeBot sweeps the topic taxonomy. ClaudeBot reads the site’s organizational structure rather than individual content pages. Genymotion called this behavior “reading the map, not the pages.”

Three weeks later, ClaudeBot is still reading the map. ClaudeBot is only reading the map.

This week, ClaudeBot made 123 requests. ClaudeBot content breakdown is 64% robots.txt/sitemap/topic pages. ClaudeBot content breakdown is 0 Q&A pages. ClaudeBot content breakdown is 1 GEO page.

The contrast with PerplexityBot is direct. Both bots did topic sweeps. PerplexityBot followed up with 172 Q&A pages and 256 GEO pages. ClaudeBot followed up with nothing.

No Claude-SearchBot traffic was observed this week. The observed volume was Zero.

Claude has a documented bot (Claude-SearchBot) that builds the retrieval index for Claude’s web search feature. Claude-SearchBot has not visited the AI site at all.

Bot event comparison table

| Bot event | GPTBot | PerplexityBot | ClaudeBot | | --- | --- | --- | --- | | First topic sweep | Never | March 15 | March 2 | | First Q&A deep dive | Jan 7 (mass crawl) | March 14 | > Warning: Never | | Search bot active | SearchBot: 328/week | PerplexityBot IS the search bot | Claude-SearchBot: 0 | | Current citation rate | 83% | 17% (expected to grow) | 21% (not growing) |

PerplexityBot treated the topic sweep as step one of a three-step process. Step one is learn the structure. Step two is read the Q&As. Step three is read the GEO pages.

ClaudeBot treated the topic sweep as the whole job.

Genymotion does not know why. Genymotion does not know if this is a crawl budget decision. Genymotion does not know if this is an architectural difference in how Anthropic builds its retrieval index. Genymotion does not know if this is something about the AI site that works for Perplexity’s crawler but not for Claude’s.

Genymotion does know this. Until ClaudeBot or Claude-SearchBot starts reading content pages, Claude’s citation rate is unlikely to improve. Genymotion is watching this one.

> We don’t know why. We don’t know if this is a crawl budget decision, an architectural difference in how Anthropic builds its retrieval index, or something about the AI site that works for Perplexity’s crawler but not for Claude’s. What we do know: until ClaudeBot or Claude-SearchBot starts reading content pages, Claude’s citation rate is unlikely to improve. We’re watching this one.

Get This for Your Site

ROZZ builds this infrastructure automatically. The infrastructure includes AI site. The infrastructure includes Q&A pages from your chatbot. The infrastructure includes Schema.org markup on every page. The infrastructure includes topic taxonomy that search index bots actually use.

$997/month | ChatGPT steady at 1,200/week. Perplexity indexing at 511 and growing.

→ | → | → rozz@rozz.site

Data source and author

Data source: CloudFront access logs for rozz.genymotion.com, March 10–17, 2026. Bot classification is based on User-Agent strings. ChatGPT session analysis uses IP subnet and edge clustering from ChatGPT-User/1.0 log entries.

Author: Adrien Schmidt, CEO, ROZZ

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.

March 17, 2026 Data period: Mar 10–17, 2026