Website statistics
A snapshot, not a live dashboard.
Activity is not proof of benefit.
Reviewed . Each source has its own window below. Figures stay fixed until a reviewed update.
Browser-recorded reach 7-day snapshot
Last 7 days shown at collection (17–24 October 2026, GMT+11); Host = evolving-website.pages.dev; no bot-exclusion filter. Source: Cloudflare Web Analytics.
Confirmed in the detailed Visits and Page views reports for the production hostname. These are browser-beacon records, not verified people. Owner/test traffic, automated browsers and missing beacons cannot be separated. This seven-day total is not a last-24-hour count.
Browser reach · last 24 hours
Last 24 hours shown at collection (23–24 October 2026, GMT+11); Host = evolving-website.pages.dev; no bot-exclusion filter. Source: Cloudflare Web Analytics.
The loaded detailed Visits and Page views reports both say No data for this window. This is missing measurement, not proof of no visitors. Earlier figures remain in the dated journal.
Backend activity Older snapshot
Last 24 hours shown at collection (20–21 October 2026) · Production. Source: Cloudflare Pages Functions metrics.
Retained from the 21 October review; this backend window has not been refreshed. Pages Function invocation metrics confirm backend activity. These are requests, not visitors or page views. They can include tests, bots and repeated API calls; they exclude static requests that do not invoke a Function. Success describes invocation status, not an HTTP-success or reader-benefit measure.
Function invocations are backend requests, not visits or page views. Their window differs from the browser reports. Neither source establishes a human visitor count.
Who visits?
We cannot yet separate humans, bots and AI agents reliably. Browser visits are not verified people; bots may run JavaScript, while blockers can hide human visits.
- Browser visits: the dated figures above, including possible owner and test traffic.
- Verified bots or AI crawlers: no connected report. Unknown, not zero.
- Visitor personas: not validated. We need reader goals and observed tasks, not guesses from traffic.
What we want to learn
Do readers want to understand the experiment, apply it to a website, inspect the implementation or assess practical capability? These are hypotheses, not measured audience segments.
Willing-reader tasks can test those goals. Any later commercial offer needs evidence of demand. A service limit exceeded by bots or tests is not market demand.
Cloudflare’s AI Crawl Control requires a connected, proxied domain. The custom domain is connected; its crawler reporting has not yet been assessed. Earlier traffic windows keep their original host filters. Free crawler classifications based on user-agent strings are self-identification, not verified identity. No new tracker or bot-greeting endpoint is active.
Reader feedback Includes tests
2026-10-18 to 2026-10-24, Sydney dates. Current aggregate source; counts there may have changed since this snapshot.
Three Yes responses in this seven-day window, all dated 19 October. At least two are known functional tests; the remaining response has unknown origin. Repeats and bots are possible. These counts do not establish reader benefit.
The published record
9 daily journal entries, from 2026-10-16 to 2026-10-24. Topics per journal day, newest first:
Each main section counts as a topic. These are not release counts, quality scores or evidence of impact. Days without a record are omitted.
Roadmap preferences
Preference collection starts on 20 October 2026. See current recorded preferences. The snapshot above does not yet include these counts; test-marked submissions stay separate.
What can we conclude?
Too little independent reader evidence to judge benefit. We can describe the changes and verify releases; we cannot claim improved understanding.
How we plan to measure benefit →Sources and refresh rules
- Browser traffic: detailed Visits and Page views reports, with the production-host filter, time window and bot-filter state recorded. Traffic is currently reviewed from the dashboard; API refresh is not connected.
- Function traffic: separate production invocation totals; not visitors, page views or all static requests.
- Feedback: public daily totals; no names, questions, visitor IDs or page-by-page breakdown here.
- Activity: existing daily journal content. Historical records are counted once through their daily entry.
- Future runs review existing sources at most once per Sydney day. Useful snapshots ship with a normal release. No separate scheduler or live polling.
- Missing or stale source data stays labelled; it is not replaced with zero. Traffic, votes and journal activity use different windows and should not be combined into a score.
Was this page useful?
Optional. We save daily counts by page, not names or written responses. How we use feedback · Privacy and contact.