Rank trackers are reading bad Google results, worst in Australia

Pages that look right but are wrong

Rank trackers work by scraping Google's search result pages at scale, counting positions for specified keywords and feeding that data to agencies and clients as performance reports. Since 2025, Google has tightened its defences against bulk automated collection, dropping the num=100 parameter, adding goto redirects and increasing human-verification challenges. Search Engine Roundtable's Barry Schwartz thinks the changes were aimed at AI bots and scrapers rather than SEO tools; Google has not said.^[2] Rank trackers are caught either way.^[1]

The result is what rank-tracking firm Nozzle calls poisoned search engine results pages (SERPs): complete-looking pages with the wrong organic listings for a keyword. Nozzle published research on 10 October covering 27.5 million search result pages across 49 countries. As the Nozzle research team said, "Nothing on the page gives it away."^[1] A tracker reads the page, logs the positions, and files an inaccurate report with no error flag attached.

Nozzle's requests to Google began failing in large numbers on 13 September, and poisoned pages first appeared on 20 September. By 29 September, 27.0% of collected pages were poisoned, against a baseline of 0.17% between 20 August and 9 September.^[1] The peak ran across three consecutive days: 24.9% on 27 September, 26.7% on 28 September and 27.0% on 29 September. Since 1 October, rates have settled at roughly 5 to 11% daily, still far above any prior level.

Australia's exposure

Of the 49 countries Nozzle measured, Australia recorded the highest poisoned rate at 11.5%. India came second at 10.9%, Canada third at 10.5%, and the United Kingdom sat at the bottom of the table at 3.0%.^[1] The Australia-versus-UK comparison is a figure Prompt the Market computed from Nozzle's country table. An Australian agency's rank tracker is 3.8 times more likely to be reading a poisoned page than its UK counterpart, and monthly reports produced here carry a materially higher risk of being wrong.

On corrupted pages, tracked sites fell out of the top 10 on 58.6% of reads, compared with 2.7% on clean ones.^[1] A client whose site actually sits in position four could look, to a tracker pulling a poisoned page, as though it has vanished from the first page entirely.

Lily Ray, a US SEO consultant, said on 8 October that rank trackers were "too unreliable at the moment" to share findings with, Search Engine Roundtable reported.^[2]

A timing problem that made things worse

The poisoning peak on 29 September coincided with a Google spam update that SEO expert Glenn Gabe confirmed on 7 October caused genuine ranking changes for some sites. That overlap made it hard to separate real losses from tracker noise. On one site whose drop the owner confirmed, Gabe found Ahrefs picked up the fall accurately, Semrush caught it late and Sistrix did not show it at all.^[2]

By 12 October, SE Ranking had posted a status notice about "missing positions for some keywords", and Sistrix, DataForSEO, Advanced Web Ranking and Morningscore had all published notes on the problem, Search Engine Roundtable reported.^[2] Advanced Web Ranking says it now re-checks suspect results before storing a position.^[3] Google has not commented.

Nozzle's data shows 77.9% of poisoned pages had their usual results back within three days, with a median recovery time of two days. Any agency that logged a ranking fall for an Australian client between 13 September and 12 October without cross-checking Google Search Console's own position data should run that check before the next report goes out.