Google Rank Tracking Is Broken Right Now: What "Poisoned" Results Mean for Your SEO Reports

If your SEO report has shown a sudden drop, a strange spike or keywords vanishing from page one since late September, do not panic and do not start rewriting pages yet. There is a good chance your rankings have not moved at all. The tool measuring them has.

On 27 October 2026, Semrush confirmed that Google is returning inaccurate search results to some automated ranking tools, and called the problem industry-wide. Ahrefs, AccuRanker, BrightLocal, SEOmonitor, SISTRIX and others have reported the same issue. The industry has started calling these "poisoned" SERPs (search engine results pages).

This guide explains what is going on, how to tell a fake ranking drop from a real one, and what UK businesses should measure instead while the tools catch up. If you work with an SEO agency, it is also a fair test of whether they report on rankings or on revenue.

What are poisoned SERPs?

Rank trackers work by searching Google automatically, millions of times a day, for set keywords, locations and devices, then recording where each website appears. Google has never liked large-scale automated scraping of its results, and in 2026 it has stepped up its defences.

According to reporting by Search Engine Roundtable on 12 October, Nozzle (a rank tracking company) described three things Google appears to be doing to suspected scrapers:

1. Blocking large ranges of IP addresses, including big residential proxy networks.
2. Showing far more captchas.
3. Serving altered results to sessions it suspects are automated.

That third point is the poisoning. The page looks completely normal, with ten blue links and the usual features, but some or all of the results have been swapped. Semrush described two types: results that do not match the search at all, sometimes on page one, and pages that look fine at the top but contain loosely related results further down. One example Semrush gave was a search in Brazil for a specific baby stroller that returned a dental supply store's equipment page, because both products shared the Portuguese word "carrinho".

Google has not commented and has not confirmed it is deliberately serving altered results to rank trackers. Semrush and other providers have suggested it is linked to Google's efforts to restrict large-scale data collection, including collection for AI systems. Treat that as a likely explanation, not a confirmed one.

How big is the problem?

Bigger than a rounding error. The figures below come from different providers using different samples and detection methods, so they are not directly comparable, but they all point the same way.

• Ahrefs said on 22 October that between 5% and 20% of search results were affected, depending on the data provider. It filters abnormal results, requests replacements and is cleaning up past data.
• Nozzle analysed 56.1 million Google SERPs across 53 countries from 20 August to 21 October. Since 13 September it flagged 22.6% of SERPs checked on both page one and page two as poisoned, up from 0.28% in its baseline period. The worst day was 30 September at 43.5%.
• SEOmonitor reported that about 8% of its daily results were inaccurate or incomplete, and that incorrect data affected roughly 10% of tracked keywords between 9 and 13 October.
• DemandSphere flagged 6.7% of collected SERPs as suspected poisoning on 23 October.
• AccuRanker reported that 40% of its initial desktop requests on 24 October returned suspected poisoned first-page results, falling to 3.2% after its checks and below 1% the next day once it adjusted its collection.

Two details matter for business owners. First, Nozzle found 67.9% of affected SERPs went back to normal within three days, so many "drops" fix themselves. Second, it found Facebook, Instagram, YouTube and Quora turned up more often in suspected poisoned results. If your report suddenly shows social and forum pages leapfrogging you, that is a warning sign the data is bad, not that your competitors got lucky.

Why this landed at the worst possible time

The poisoning did not arrive on its own. Google released its September 2026 spam update on 24 September, which caused genuine ranking changes for many sites, up as well as down. We covered that in our guide to the Google spam update September 2026.

So since late September, two things have been happening at once: real ranking movement from the spam update, and fake ranking movement from bad tracker data. SEOmonitor made exactly this point in its help centre. If you only look at a rank tracker, you cannot tell which is which.

Add Black Friday and Christmas trading, when every lost enquiry costs more, and you have a recipe for expensive overreactions: rewriting pages that were fine, cutting content that was working, or panic-buying ads to cover a drop that never happened.

How to tell a fake ranking drop from a real one

Before you change anything on your site, run these checks.

1. Check Google Search Console first

Search Console is Google's own data about your site, based on real searches by real people. It is not scraped, so it is not affected by poisoned SERPs. Semrush's advice to customers who see unusual ranking changes is to compare their reports with Google Search Console.

Open the Performance report, set the date range to compare the last 28 days with the previous period, and filter by the page or query that "dropped". If impressions and clicks are steady, the drop is almost certainly a tracking problem. If they have fallen in line with the tracker, it is probably real. If you are new to the newer reports, our guide to the Search Console AI performance report walks through what each view shows.

2. Look at the keywords around it

Poisoning tends to hit keywords individually and briefly. Genuine algorithm changes tend to move a whole group of related pages and queries together, and the change holds. One keyword falling 30 places overnight while its neighbours stay put is a classic fake.

3. Wait three days before acting

Given Nozzle's finding that most affected SERPs returned to normal within three days, a sensible rule is not to act on a single day's ranking change. Real changes usually hold across consecutive days.

4. Check the pages that "beat" you

If the new results above you are unrelated to the search, or a sudden flood of social posts and forum threads, be suspicious. Search the keyword yourself in a private browser window from a UK location and see what a real person gets.

5. Check your leads and sales

The most honest test of all. If enquiries, calls, form fills and online sales from organic search are where you would expect for the time of year, your visibility has not collapsed, whatever the chart says. This is where measuring marketing ROI properly pays for itself.

What this means for SEO reporting

For years, plenty of SEO reports have led with a table of keyword positions. It is easy to read and it looks like progress. This episode shows why rankings should never be the headline number.

Rankings were already a rough guide. Positions change by location, device and search history, and Google's results now include AI Overviews, maps, videos and shopping results that push the classic links down the page. Since Google removed the option to load 100 results per page in 2025, trackers have also had to work harder for deeper data. Now, for a period at least, some of that data is simply wrong.

A better report puts the business numbers first:

• Organic enquiries, calls and sales, tied to the pages that drove them.
• Search Console clicks and impressions by page and by query group.
• Click-through rate on your most important commercial pages.
• Rankings as supporting context, with a note on data quality where relevant.

That is how we report at BlackFire. Our view has always been that lead quality beats lead quantity, and the same goes for rankings. A first-page position on a keyword that never brings a customer is not worth celebrating.

Should you change your rank tracker?

Not necessarily. Every major provider is dealing with the same problem, and several have now added checks. AccuRanker, for example, compares each new page one with the sites that ranked on at least two of the previous three days and re-collects the page if fewer than half of them appear. SEOmonitor and DemandSphere have added similar history-based checks, and Ahrefs is filtering and replacing abnormal results.

What is worth doing is asking your provider, or your agency, three questions:

1. What checks do you run to detect and exclude poisoned results?
2. Have you cleaned or reprocessed historical data from September and October 2026?
3. How do you flag data you are less confident in?

If nobody can answer, take the ranking charts with a pinch of salt and lean harder on Search Console.

What UK businesses should do this week

• Do not make big site changes because of a rank tracker alone. Confirm in Search Console and in your sales data first.
• Annotate your reports. Mark late September to now as a period of unreliable third-party ranking data, so nobody misreads it in six months.
• Separate the spam update from the tracking issue. If Search Console shows a genuine, sustained drop starting around 24 September, follow the steps in our spam update guide.
• Keep local SEO in view. BrightLocal said the problem mainly affected rankings below position 10, but local businesses should still check their Google Business Profile insights directly. Our Google Business Profile SEO guide covers what to watch.
• Do not forget AI search. Visibility in AI Overviews and AI Mode is not captured well by classic rank tracking anyway. See our guide to AI search optimisation.
• Hold your nerve on budget. Moving money from SEO to paid ads to "cover" a drop that is not real wastes budget at the most expensive time of year. If you are weighing the two, read SEO vs paid ads for lead generation.

FAQ

Why have my Google rankings suddenly dropped in Semrush or Ahrefs?

Since September 2026, Google has been returning inaccurate results to many automated rank trackers, and Semrush and Ahrefs have both confirmed it. A sudden drop in a third-party tool may not reflect what real searchers see. Check Google Search Console before acting.

Is Google Search Console affected by poisoned SERPs?

No. Search Console reports on real searches by real users, using Google's own data, so it is not affected by the scraped results that rank trackers collect. It is the best place to confirm whether a ranking change is real.

How long will rank tracking be inaccurate?

Nobody knows yet. Google has not commented, and providers say their checks have reduced but not eliminated bad data. AccuRanker reported further changes on 20 and 23 October. Expect some noise to continue and judge performance on Search Console, enquiries and sales.

Did the September 2026 spam update cause my ranking drop?

It might have. The spam update began on 24 September and caused genuine ranking changes for many sites. The way to tell is Search Console: a real drop shows up as a sustained fall in impressions and clicks, while a tracking error does not.

Next step

If your SEO reports have been swinging around and you are not sure what is real, we can check your Search Console data, separate tracking noise from genuine changes and tell you plainly where you stand. Book a call or email hello@blackfiremarketing.co.uk, and see how our SEO service reports on leads and sales, not vanity charts. You can also start from our home page.