AI Search

How to read crawler share by billing region

Read Cloudflare’s regional Googlebot shares with the right geography and denominator, then choose the agents and pages to inspect in your own crawl records.

Six trays of green and gray page tokens with a magnifying glass over one tray

Key takeaways

  • Cloudflare groups the sites in this benchmark by their owners’ billing region. [1]
  • Googlebot’s share of measured crawler traffic ranged from 35.8% to 54.2% across the six October 2025 cohorts. [1]
  • Each regional percentage uses its own denominator. Comparing request volumes requires the underlying counts. [1]
  • When a share changes on your site, inspect the agent’s count, total recorded crawler requests and affected pages before choosing an action.

Use a regional crawler share to frame a question for your own records. Check whose geography the benchmark describes, which traffic forms its denominator, and whether the agent is reaching the pages you care about. Then use the recorded counts and responses on your site to decide where to investigate.

Cloudflare’s October 2025 regional analysis provides a useful worked example. We examine the Googlebot entry for each of its six billing regions, calculate the remaining share and show how a publisher can turn that comparison into a focused page review. The observations describe a historical month on Cloudflare’s network. [1]

Start with the geography behind the label

Cloudflare assigned the crawled sites to regions using the billing addresses of the customers who owned them. Googlebot’s percentage in each cohort is its share of the crawler traffic measured for that cohort. A company’s billing address can differ from its audience’s location, its hosting location and the origin of a crawler request. Choose the geographic definition that fits your question before comparing reports. [1]

For example, an editorial team managing a multilingual help center might want to know whether its French documentation receives enough attention from relevant crawlers to merit a closer review. The most direct scope for that question is the French documentation URL group. The team can inspect requests to those pages, even if the business that owns the site is billed in another region. The regional chart gives context for the broader mix of activity. The page records identify what happened to the content under review.

Write the scope in one sentence at the start of the review. Include the website or page group, dates, recorded agents and responses. If the concern is tied to a visitor market, record how that market was identified. This helps the team choose comparable evidence when it returns to the same question next month.

Read the full regional comparison

The chart includes the six published Googlebot shares. Ansvra calculated the other segment in each bar by subtracting the reported percentage from 100%. Every bar has its own regional denominator. Cloudflare published these observations in The 2025 Cloudflare Radar Year in Review, under “Crawling by geographic region.” [1]

Googlebot’s share by customer billing region

Measured crawler traffic in six Cloudflare customer billing regions, October 2025.

1007550250N. Am.S. Am.EuropeAsiaAfricaOceaniaCustomer billing region
  • Googlebot (reported)
  • Other crawlers (calculated)
Full data table
Googlebot’s share by customer billing region · % of measured crawler traffic
Customer billing regionGooglebot (reported)Other crawlers (calculated)
N. Am.45.554.5
S. Am.44.255.8
Europe48.651.4
Asia3961
Africa35.864.2
Oceania54.245.8

Cloudflare reported the Googlebot percentages. Ansvra calculated the remaining share as 100% minus each published value. Each bar uses its own regional denominator. Request counts were not supplied. N. Am. and S. Am. denote North America and South America. [1]

The highest displayed share is 54.2% in Oceania, and the lowest is 35.8% in Africa. Their difference is 18.4 percentage points. Oceania is the only cohort where Googlebot accounts for more than half of the measured crawler traffic. In the other five, Googlebot is the leading single agent in the published table, while the remaining measured crawlers collectively account for more than half. [1]

A high share describes the composition of a cohort. To compare total Googlebot activity between two regions, you would also need each cohort’s total crawler activity. The selected table supplies percentages without those totals or site counts. Preserve that gap when reusing the figures. [1]

Put the counts beside a changing share

The same composition issue appears in a website’s own records. An agent’s share depends on both its recorded requests and the total eligible crawler requests. A change in another agent’s activity can move the first agent’s percentage even when it makes fewer requests.

Consider the following arithmetic illustration. These invented counts explain the calculation and are separate from the Cloudflare observations.

Illustrative windowAgent requestsAll eligible crawler requestsAgent share
First window4010040%
Second window306050%

In the second window, the agent’s share rises by 10 percentage points even though it makes 10 fewer recorded requests. Other recorded crawler requests fall from 60 to 30. An analyst who saw only the percentages would miss the drop in the agent’s count and the larger change in the rest of the recorded activity.

Begin by checking that both windows have comparable coverage. A recording connection added halfway through the first period creates a different comparison from two complete periods. A revised rule for grouping agents can also change which requests contribute to a total. Record the change, then use a common scope where the evidence supports one.

If a count is unavailable, leave it unavailable. If total eligible crawler requests are zero, the share is undefined. Keeping these states visible helps the next reviewer distinguish an absence of recorded activity from an incomplete observation window.

A magnifier over page cards beside two stacks of crawler tokens and a separate bookmarked card
Move from the overall mix to the pages and evidence relevant to the review.

Choose the agent and pages worth inspecting

Start with a reason to investigate, such as fewer recorded requests to a new documentation section or a change in responses for an important product page. Use the overall agent share to orient the review, then move to the URLs that matter for that question. A small share across the website can still include activity on a critical page.

Use a regional benchmark to start a site review

Five checks for moving from an external request share to your own crawl records and answer evidence.

  1. Check scope

    The chart groups requests by the billing address of the customer that owns the crawled site. Treat the shares as historical network context. [1]

    Next: Open records

  2. Open records

    Choose one website or section and a complete date range. Write down which agents and responses your records cover. [1]

    Next: Read counts

  3. Read counts

    For each relevant agent, record the observed request count and the total eligible crawler requests. Check both before interpreting a change in share. [1]

    Next: Check pages

  4. Check pages

    Inspect which page groups and response types changed for the agent. Give priority to URLs that matter to your audience or the review. [1]

    Next: Review answers

  5. Review answers

    Use the saved answers for the relevant questions to see which pages were cited. Keep the answer evidence separate from the crawl review. [1]

Ansvra’s suggested review workflow. Cloudflare is cited for the benchmark scope and definition. [1]

For each agent under review, look at its count alongside the total eligible crawler requests in the same window. Then break the agent’s activity down by the page groups your question concerns. This can reveal whether a change is broad or concentrated on a small part of the site. Keep the response information next to the affected URLs so the team knows where an access review may help.

Suppose the site’s overall crawler mix is steady but a newly launched guide has few recorded requests. The next check is whether the relevant agent requested that guide during a complete window and what responses were recorded. Suppose instead that the same agent’s activity changed across many established pages. A wider review of the recording coverage and page responses would better fit that observation. Neither scenario can be chosen from a regional percentage alone.

In Ansvra’s AI Agent Analyzer, teams with a supported connected site can review recorded crawler activity, pages and responses. Use the dates covered by those records and identify the URL group before drawing a conclusion about the change. AI Agent Analyzer

Use answer evidence for the citation question

A crawl review helps you inspect recorded requests and responses. When the business question is whether a product or guide appears in AI answers, open the saved answers for the relevant prompts as well. Record the cited URLs and the conditions of the answers you compare.

For a documentation review, that might mean checking whether the help page is referenced for a question it directly answers. If another page appears, read the answer and the cited source before deciding what to improve. A page can be relevant to an editorial priority even when the request totals provide little detail about its use in an answer.

Citation Source Analysis helps teams inspect available references and recurring cited pages in saved answers. Bring that evidence into the same review when the question concerns citations, and retain the prompt and answer context alongside each URL. Citation Source Analysis

Record a decision the team can revisit

Close the investigation with the narrowest action supported by what you observed. If the change is concentrated in a page group with different responses, assign a review of those URLs. If it follows a change in the recording window or agent grouping, first establish a comparable period. If the concern arose from a saved answer, keep the cited page and prompt attached to the content task.

The handoff should say which agent and URLs were reviewed, which dates were covered, what the recorded counts showed and what remains to be checked. Name the evidence needed for the next decision. A page response review needs different material from a content review, and an unexplained change in a site-wide share may require a breakdown of other recorded agents.

On the next review date, return to the same scope. Compare counts and page responses, then inspect relevant saved answers if citations are part of the question. That gives the team a repeatable way to evaluate the action it took.

Method and limits of this analysis

We extracted the Googlebot cell from each regional row in Cloudflare’s October 2025 table. The complete scoped dataset has six reported values. The chart’s remaining shares and the 18.4-point spread are Ansvra calculations based on the published one-decimal percentages. The full chart table contains every value used in the comparison. The illustrative count table uses invented values solely to explain share arithmetic. [1]

The report does not provide the underlying counts or enough regional sampling detail to reproduce this table independently. Its billing cohorts describe aggregate activity across observed customer sites. Use the displayed shares to understand that historical composition, and use your own complete records to investigate your site. [1]

Summary

Read a regional crawler share together with its geography, date and denominator. For a site review, choose the agent and page group that match your question, inspect both request counts and responses, and record the next check. When the question concerns citations, bring the relevant saved answers into the review.

Sources

  1. The 2025 Cloudflare Radar Year in Review: The rise of AI, post-quantum, and record-breaking DDoS attacksAccessed

Discussion

Join the discussion

Start the conversation

Share a question or tell us what worked for you.

Comments appear after review. How we handle comments

The Ansvra journal

Keep reading

Browse all articles

Ansvra

A clearer next step.

MEET YOUR AI GUIDE

A little clarity.
A clearer next move.

Ask about Ansvra, AI visibility, or your next step with content.

Opening chat…