Credited — the measurement series
Do AI assistants send anyone to my site?
A measurement I run on a fixed schedule and publish whichever way it goes: ten questions, four AI assistants, and the crawler log underneath. Round 0 scored 31 out of 80. Round 1 scored 25 — and the knowledge base went from zero citations to seven.
This page is published in English only.
Analytics tells you who arrived. It never tells you who was recommended and did not come. So I measure it by hand: the same ten questions, to ChatGPT, Claude, Perplexity and Google, in a fresh session with no account and no memory — plus the crawler log, which says whether those assistants ever fetched the pages at all.
Every block below is dated and stays as it was measured. Nothing on this page gets rewritten later; new rounds are appended underneath. The full method, and the reasoning for each number, is in the article that started this — and each round gets a video in the Credited series.
Round 0 — 28 & 29 July 2026
What Round 0 measured
Two instruments, because one of them cannot answer on its own. The first asks whether the assistants send anyone here. The second asks whether they ever opened the pages at all — because a zero in the first table means one of two opposite things, and only the crawler log tells you which.
Do they link me?
The ten fixed questions, asked of four assistants in a fresh session with no account and no memory, scored by hand. Two points for a link to this site, one if the work was in the answer but the link went elsewhere, zero if absent.
- 2 — it linked this site
- 1 — my work, someone else's link
- 0 — not there at all
Eleven of the forty answers scored one. The work is in the answer and the credit is somewhere else — a YouTube video, or a free Webflow project living on Webflow's domain. That is a different problem from being unknown, and it has a different fix.
Of those 31 points, the knowledge base scored zero. Not one of its 42 pages was cited, by anyone, once. That is the reason for measuring before announcing: every point on the board came from something that existed beforehand, so the “before” is clean.
Did they read the pages at all?
Requests from AI crawlers to the 42 pages of the knowledge base, on the day it was published and on the day after the announcement put the first links to it on the internet.
/kb/, in one day - 28 July — the day it went live
- 29 July — the day after the announcement
0 — 47 requests
The six named crawlers add up to exactly 117, so OpenAI's zero is a real zero — not a vendor hidden inside a “+N more” chip.
Anthropic read essentially the whole bundle in one hour. Around 44 of its 47 requests fall inside a single hour on 29 July. The bundle is 42 pages.
Even Huawei and Apple fetched a page. OpenAI fetched none — not on the day it was published, not after the announcement.
Source: Cloudflare AI Crawl Control, filtered to /kb/gsap-webflow/*,
read and recorded by hand on 2026-07-28 and 2026-07-29. This chart is drawn from
those recorded numbers — it is not a screenshot of the dashboard.
| Crawler | 28 Jul | 29 Jul |
|---|---|---|
| Anthropic — ClaudeBot | 0 | 47 |
| Amazon — Amazonbot | 3 | 31 |
| Google — Googlebot | 8 | 20 |
| Perplexity — PerplexityBot | 1 | 17 |
| Huawei — PetalBot | 0 | 1 |
| Apple — Applebot | 0 | 1 |
| OpenAI — GPTBot · ChatGPT-User | 0 | 0 |
| Total | 12 | 117 |
Update — 1 August 2026
OpenAI is no longer at zero
Three days after the announcement put the first links to these pages on the internet, OAI-SearchBot — the crawler behind search inside ChatGPT — fetched two pages, 16.04 kB.
Two pages out of forty-two, and a different agent from GPTBot. Cloudflare's view here only goes back twenty-four hours, so this is the first time I have seen it — not necessarily the first time it happened. Read on 2026-08-02, ~11:15 JST.
Round 1 — 28 August 2026
What Round 1 measured
Same ten questions, same four assistants, same scoring. Two numbers moved in opposite directions on the same day: the total fell, and the knowledge base was cited for the first time.
- 2 — it linked a knowledge-base page
- 2 — it linked this site, another page
- 1 — my work, someone else's link
- 0 — not there at all
Perplexity scored 14 again — and almost nothing underneath is the same. In Round 0 its points came from Academy lessons and YouTube videos. Six of its seven scoring answers now link the knowledge base instead. Three cells that used to be a 1 — the work named, the credit elsewhere — are now a 2.
ChatGPT fell furthest, and Claude never touched the knowledge base. Claude's two scoring answers both link the old Supasaito addresses the Academy used before it moved here; they still resolve, because the redirects hold.
7 of the forty answers cited nobody at all. Not “cited someone else” — no sources, no panel, nothing. The scoring writes that the same way it writes every other zero, and it means the opposite thing: a contest lost can be won back, and a question the assistant answers from itself has nothing to win.
How much is a reading worth?
I reread two columns within a day of the round, with nothing on the site changing in between. They moved — in both directions.
- ChatGPT 5 1 4 28–30 Aug
- Perplexity 14 18 28–29 Aug
Four points, either way, inside a day. So every figure on this page is a reading rather than a fact, and the honest form of any of them is a range. What survives that is the size of the change underneath: a wobble of two citations does not turn zero into six.
Were they reading it?
The same crawler check, same 24-hour window, same /kb/ filter —
against Round 0's reading a month earlier.
OpenAI went from never having fetched a page to six of them — and cited none. Its crawlers opened the exact pages that answer the questions I asked it that day. The score for ChatGPT's knowledge-base column is still zero. Being read and being cited are not the same event, and this is the first round where both instruments can show it inside one day.
Anthropic went the other way: 47 requests to zero. Its crawler visited the site 26 times that day without opening a single knowledge-base page. Same zero on the scoreboard as ChatGPT's, opposite cause underneath — and only the crawler log tells them apart.
And nobody arrived. Across 148 distinct paths the AI crawlers touched that day, referral traffic to this site was zero. Including Perplexity's six citations, which are the best result in the whole experiment.
Next
The quiet-period reread — 6 September 2026
Asking an assistant about your own pages sends its crawlers to those pages. So the suspicion Round 1 raised is that the measurement fed the thing it was measuring — and the test is a week of silence, then the same ten questions to ChatGPT and Perplexity on the same day. If Perplexity falls back toward 14, every month-over-month comparison in this field has a problem. Published either way.
Round 2 stays where it was promised: three months from Round 0, so late October.