The 662-Page Indexing Report: Two New States, Position 21
Search Console labels and behavior change over time. Every figure here is a snapshot from October 4–5, 2026 — verify against current Google documentation before acting on it.
Last week I published what week one actually looks like on a 662-page static site: one indexed homepage, everything else “URL is unknown to Google”, and a promise to run the full-batch inspection in week two. I ran it a day early, on day 7. Here are the final week-one numbers: all 662 URLs inspected, zero state flips on the big site — and the indexing states I was waiting for showed up the same week, on the family’s smallest site instead. Plus the first search data worth reading: 92 impressions and a position-21 head term.
The full batch: 662 inspected, and the queue did not move
The full inspection I deferred from launch day happened on day 7 (October 4, 2026), once the results could rank pages against each other. Here is the complete distribution:
| State | Count | Share |
|---|---|---|
| Submitted and indexed | 1 | 0.2% |
| Discovered – currently not indexed | 0 | 0 |
| Crawled – currently not indexed | 0 | 0 |
| URL is unknown to Google | 661 | 99.8% |
The one indexed page is still the homepage, last crawled October 1, and its referring URLs are still a link from my blog’s tools page plus the site itself. Meanwhile the sitemap grew from 568 to 662 URLs during the week and processed with zero errors. Six days after my first post, the queue looks exactly the same at full resolution as it did in a 12-page sample: this is what a healthy week one on a new programmatic site looks like, and the honest headline is that there is no headline yet.
One practical note for anyone running the same play: 662 individual URL Inspection calls took about 80 minutes at a polite per-request delay, and roughly one in four requests failed with a transient SSL error and needed retrying. Batch with resume support, or you will re-spend that quota every time something interrupts the run.
The states showed up — on the 48-page site, not the 662-page one
The week’s most informative data came from the smallest site in the family, a 48-page work-hours calculator site called TallyPunch that is three weeks older than the word-list site. A seven-URL scan there captured the family’s first two non-terminal indexing states:
| URL | State (Oct 4) |
|---|---|
| / | Indexed |
| /about/ | Indexed |
| /biweekly-time-card/ | Indexed |
| /what-is-time-and-a-half/ | Indexed |
| /time-card-calculator/ | Indexed |
| /how-many-hours-is-12pm-to-8pm/ | Discovered – currently not indexed |
| /how-many-hours-is-9am-to-5pm/ | Crawled – currently not indexed |
Same family, same infrastructure, same sitemap-and-internal-links launch pattern — opposite trajectories. My working hypothesis, and it is only a hypothesis: a 662-page sitemap lands in Google’s queue as one big batch that gets paced, while a 48-page site gets crawled to completion in days. Small sites get depth first; big small sites wait in line longer. If the week-two rerun on the big site shows the first flips, I will take that as support.
The checklist-driven responses are deliberately boring. The Discovered page gets an internal-link priority check (it currently has no referring URLs) and then patience. The Crawled page — Google fetched it on September 28 and declined to index it — gets two weeks of observation with zero changes, until October 18, before any attribution. Changing a page during its observation window destroys the experiment.
The first query data worth reading
Indexing states tell you what Google did. Query data tells you what searchers did. Until this week, every site in the family sat under my 20-impression floor: below that, I write “insufficient sample” and touch nothing. This week the smallest site crossed it.
The 48-page site recorded 92 impressions across 50 query rows in the measurement window (September 22 to October 1; the API’s rolling window lags by a few days). Zero clicks, so click-through rate is still meaningless. What matters is position:
- Head term: “30 minute lunch break calculator” — 4 impressions at position 21. A calculator page is one page away from page one of Google for its exact keyword, three weeks after launch.
- Same cluster, hyphenated variant: “30-minute lunch break calculator” — 4 impressions at position 38.8, a reminder that near-duplicate queries rank independently.
- Mid-tail: time-card and hours-to-decimal terms each show 1–2 impressions at positions 65–86 — normal new-site depth, nothing to act on.
One page at position 21 is not a trend. It is the family’s first discriminating signal: proof that a brand-new domain with zero backlinks can get within sight of page one on a specific, answer-shaped query. The next checkpoint is whether it enters the top 10, and that is a week-two question, not a week-one one.
What I changed this week: nothing
The checklist I wrote on launch day said “unknown = normal, check again in two weeks”, and the full-batch data says the same thing with 662 rows instead of 12. The complete action list from this inspection:
- Put the Crawled-state page on a watchlist with a hard date — no edits before October 18.
- Recheck the Discovered-state page at the next inspection; add an internal link entrance only if it stalls.
- Rerun the 662-page batch next week, same caliber, and compare flips.
- Keep interpreting nothing under 20 impressions.
Boring is the point. New-site indexing is a queue, and the data so far rewards the people who wait in it without fidgeting.
A meta footnote
The week-one post I promised this follow-up from is, as I write this on October 5, itself “Discovered – currently not indexed” — 38 hours after publication. The post about that state now lives in that state. Both are queue positions, not verdicts; that was the thesis, and apparently it applies to the thesis too.
This article was created with the help of AI. AI was not used to write the content; it assisted only with translation and grammar checks.