Google Is Feeding Rank Trackers Fake Results. 8% of SERPs Came Back Poisoned.
There is a category of SEO data you have probably never questioned: the ranking numbers in your dashboard. Here is a sentence from a vendor deep dive published this month that should change that.
“Since September 2026, Google has been poisoning the SERPs it serves rank trackers: complete, normal-looking pages whose results have been swapped out. Nothing on the page gives it away. Each keyword’s own history does.”
That comes from Derek Perkins at Nozzle, who has been unusually transparent about what the tooling industry is dealing with. His measurement: 8.0% of SERPs came back poisoned.
What Google is doing, per the vendor watching it
This is not a targeted attack on rank trackers. It is the by-product of a fight between Google and scrapers — the AI bots harvesting search results, and commercial SERP APIs. Google started testing goto redirects in search results back in July to identify scraper traffic, and has been testing human-verification methods at higher-than-usual rates.
The three-part approach Nozzle describes:
- Very wide IP blocking, aimed at the major residential proxy networks that the tracking tools depend on.
- A significant increase in captcha challenges.
- Injecting bad data into suspect SERPs, so that when a tool does successfully scrape, it receives wrong answers.
The third one is the one that should worry you, because it is invisible at the point of use.
Why poisoned pages are worse than blocked ones
A blocked request is loud. Your tool reports an error, you see a gap in the chart, you know you are missing data.
A poisoned page is silent. It renders normally. It contains results. They are simply not the results a human would see. And according to the analysis, it passes within a few days — so by the time you notice, there may be nothing left to notice.
How to spot it
The identifying pattern Nozzle describes gives you three tests:
| Signal | What it looks like |
|---|---|
| Discontinuity against history | The page looks fine, but the keyword’s own trend does not |
| Social and forums filling gaps | Ordinary results replaced by a cluster of social and forum links |
| Tracked sites vanish together | Everything you track falls out of the top ten at once |
| Self-resolving | The anomaly is gone in a few days |
The third signal is the most practical early warning. Your tracked set falling out of the top ten simultaneously is not a ranking event; it is a data event. Real ranking losses are uneven.
What to trust instead
This is a good moment to notice how much of your reporting rests on scraped data, and to put more weight on things that come from a source you control or can verify directly.
Search Console, which is Google’s own data, is imperfect — the reports are sometimes delayed and the links report swings — but it is not fabricated.
Your own dated record of placements. You can open a URL and see whether it exists, whether it still mentions you, and whether the page is indexed. Nobody can poison that check.
A stable, small set of queries checked by hand. Ten queries, once a week, checked in a browser, will not give you a dashboard. It will tell you whether the dashboard is lying.
That first one — the record — is the layer most teams never build, and it is the one that survives every change in the measurement landscape. It is what a service such as https://backlinksubmit.org is oriented around: submit your product URL, get matched to relevant directories from a handpicked list, have the team submit manually, and watch a console that shows which platforms have been submitted and how each is progressing — a record that does not depend on anyone’s ability to scrape a search page.
The bigger lesson
Rank tracking has been treated as ground truth for fifteen years. What this episode reveals is that it is a third-party observation of someone else’s page, obtained against that someone’s wishes, and now actively contested.
That does not make it useless. It makes it a single instrument that occasionally lies, which is exactly why you want a second instrument that measures something else — your placements, your indexation, your branded demand — rather than a second dashboard measuring the same contested thing.
FAQ
Is Google deliberately sabotaging rank tracking tools?
Not as a first-order goal. The blocking and poisoning appear to be anti-scraping measures aimed at AI bots and SERP APIs, and rank trackers are collateral. The effect on their data is the same either way.
How much of the data is affected?
The published figure from one vendor’s analysis is 8.0% of SERPs sampled coming back poisoned, on top of requests blocked outright.
What should I do if my rankings suddenly collapse?
Check whether the change is uneven. If every tracked keyword moved the same way on the same day, suspect the data. Confirm with a manual browser check on a handful of queries before acting on it.
Does this affect Search Console?
No. Search Console is Google’s own reporting, which is why it is worth keeping as your primary source even when it is delayed or incomplete.
Share Article
