ToolsThatRank / measured data
Replication, same properties
We Re-measured the First 30 Days on 49 Measurable New Pages. The Clock Did Not Slow Down.
Our first study measured 1,185 new pages and found 17.8% got their first Google impression within 30 days. Numbers like that age fast, so we re-ran the same measurement on every page this fleet published after that study closed: 49 measurable pages (74 in scope, 25 excluded as pre-existing re-adds), published 2026-08-26 to 2026-09-18, watched daily through 2026-09-25. By day 24, 19.1% had been shown (Kaplan-Meier), with the raw empirical share at 18.4% (9 of 49 pages). The honest reading: still slow, still silence for most pages in month one, and no evidence the clock has sped up.
Verified 2026-09-28 against the Search Console searchAnalytics endpoint. Fresh pull: 48,066 page-day rows across 49 properties, window 2026-08-26 to 2026-09-25.
Check a page age
Where a page age sits on the re-measured curve
Pick your page's publish date. The tool reads the day count against the re-measured distribution from these 49 pages and says how it compares to the original study. It reads nothing from your account.
Step lookup on the published curve, capped at day 30 where this study's observation is thickest. The tool says so instead of guessing beyond the data.
The re-measurement
What the fresh cohort shows
We committed to one rule before looking: same join, same estimand, same Kaplan-Meier machinery as the first study, on a fresh GSC pull made for this page. Every filter is counted. Of 98 pages newly added to git since 2026-08-26, 24 sat on domains with no Search Console property in the account (mostly an off-fleet links site). 25 further URL matches turned out to be pre-existing pages re-added by git history; impressions dated before the first commit exposed them and they were dropped rather than counted as instant rankings. That leaves 49 pages in the measured cohort.
| PAGE AGE | FRESH COHORT (49 PAGES) | ORIGINAL STUDY (1,185 PAGES) |
|---|---|---|
| day 7 | 12.2% | 4.1% |
| day 14 | 14.5% | 9.0% |
| day 24 (fresh window limit) | 19.1% | 14.1% (read from the original curve between its printed day-14 and day-30 points; the original study does not print day 24) |
| day 30 | beyond this study's window | 17.8% |
The fresh window closes at day 24 for the youngest pages and day 30 for the oldest (published 2026-08-26, data through 2026-09-25). We do not quote past it.
Reading it
What the comparison does and does not say
The fresh cohort's day-7 share (12.2%) is higher than the original study's (4.1%). That is a small sample, 9 events in the first week, and the confidence interval is wide enough to cover the original figure. What we can say: nothing in the fresh data shows Google slowing down further, and nothing shows a step change either. The median wait in the fresh cohort had not been reached by day 24, consistent with the original median of 78 days.
One honest anomaly. 25 of the 98 newly committed pages had impressions dated before their git first-commit date. These are old pages whose git history was re-created, not new pages that Google ranked instantly. The first study caught the same class of bug by checking for negative durations, and we applied the same check here. Dropping them moved the day-24 Kaplan-Meier share from 12.4% to 19.1%, the largest single judgment call in this page, and we would rather show both numbers than hide one.
Method
How we measured this
Fresh GSC pull
One new harvest on 2026-09-28: page and date rows for every property in the account, 48,066 rows across 49 properties, window 2026-08-26 to 2026-09-25, dataState final. Two properties 403'd and are excluded.
Publish dates from git
Same harvester as the first study: first git add per content HTML page across the 43 fleet repos, restricted to commits after the first study's data closed. 98 pages found.
Same survival machinery
Kaplan-Meier with right-censoring at each property's last data date, identical to the first study. Pages with impressions before their first commit are excluded as pre-existing, and the exclusion count is published above.
What is not here
No click or top-10 curves for the fresh cohort: with 11 events, click and rank milestones would be single-digit counts dressed up as curves. We show the one milestone with enough support and say so.
Limits
What this page cannot tell you
Seventy-four pages is a replication of a small slice, not an independent population: same operator, same template-built young sites, same month of the year is not even guaranteed across cohorts. The fresh window is 24 to 30 days, so every milestone past day 30 is unobservable by construction. The day-24 comparison mixes cohort ages: the KM estimate handles that censoring correctly, but the empirical 18.4% is the rawer number if you distrust the estimator.
No ranking, traffic, or revenue outcome is promised.
The operator sells the software described here and owns every site in the dataset.
Search Console is authenticated, so no third party can verify these readings from outside.
Figures are read on the stamped date and change as Search Console reprocesses data.
Related reading. the original time-to-rank study, CTR by position, measured below position 10, and the zero click pages census, all measured on the same properties.