Google Results Under Scrutiny

SEO platforms are introducing checks to identify unusual Google search results, prevent inaccurate rankings from reaching reports and review data that may have been affected. However, their approaches differ, particularly when validating results, displaying temporary ranking changes and correcting historical records.

Nozzle Reports a Rise in Unusual Google Results

On 10 October, rank-tracking company Nozzle published research showing that Google search results pages were sometimes missing many or even all of the websites normally found for a particular query. The report was updated on 21 October, following further analysis.

Nozzle classifies a first-page result as affected when fewer than half of the websites normally appearing on page one are found anywhere in the collected results. This assessment does not require those sites to reappear in a subsequent check.

Using this method, Nozzle found that 8.7% of the first pages it assessed were affected between 13 September and 21 October. By comparison, the figure was 0.17% between 20 August and 9 September.

The company believes the issue is linked to Google’s response to AI scraping, with rank-tracking providers also being affected.

How SEO Platforms Validate Search Results

Different SEO providers use their own methods to identify unusual results before saving ranking data.

AccuRanker discards shortened Google results pages and retrieves the page again through a separate browser session. It then compares the new first-page results with websites that appeared on page one for the same keyword on at least two of the previous three days.

If fewer than half of those sites appear, AccuRanker fetches the results again. When the second retrieval shows the same new set of websites, the company treats the change as genuine.

However, AccuRanker says it cannot run this historical comparison for approximately one in every 20 desktop results and one in every 12 mobile results because there is insufficient ranking history.

SEOmonitor takes a different approach by comparing incoming results with the recent history of each keyword. The company estimates that around 8% of the Google results it receives each day are incomplete or incorrect and are retrieved again before being stored. It introduced an updated validation check on 13 October, although it has not specified the period covered by its 8% estimate.

Ahrefs also uses filters to detect unusual results, reject them and request replacements. The company estimates that between 5% and 20% of search results have been affected in recent weeks, depending on the data provider. However, it has not clarified whether this estimate refers to results before or after its filtering process.

BrightLocal is working with existing data suppliers and adding further providers to reduce its reliance on any single source. Its chief executive, Myles Anderson, said there is no single solution that completely resolves the issue. The company’s published explanation does not detail checks applied to every individual result.

What Users See While Rankings Are Being Checked

The ranking shown in an SEO platform may depend on its validation procedures and how it handles results that have not yet been confirmed.

SEOmonitor says it may retain a previously recorded ranking for up to two daily updates when it detects an unusually large decline, such as a drop from position four to position 14 or from the top 50 to beyond position 99. This gives the platform time to establish whether the change persists.

The company also provides a recrawling option and a Chrome extension that allows users to compare rankings.

AccuRanker performs an additional retrieval when its historical checks indicate that most of the usual websites have disappeared from the first page. This check takes place before the result reaches the customer’s dashboard, although the company has not explained precisely what users see while the review is underway.

Users can also request a keyword refresh within AccuRanker, prompting a new retrieval through a separate browser session.

How Providers Are Reviewing Historical Ranking Data

Some companies have also addressed results that were recorded before their updated checks were introduced.

SEOmonitor says it reprocessed keywords affected by incorrect results between 25 and 28 September. It also reports that its system prevented erroneous data from entering the platform until 8 October.

From around 9 October, however, many affected results no longer displayed the indicators used by its quality-control checks. SEOmonitor estimates that approximately 10% of monitored keywords experienced incorrect results between 9 and 13 October, although most movements in its Visibility metric were reportedly small.

The company has not specified which historical daily figures were subsequently revised.

AccuRanker uses what it calls a sandwich check to identify suspicious historical changes. The method looks for days when fewer than half of the websites appearing on page one match those found on the previous and following days.

The company tested a fixed sample of one in every 200 keywords. Its results showed that 30.2% of desktop keyword-days were flagged on 25 September, compared with 0.58% on 17 October. AccuRanker has not confirmed whether flagged historical data will be amended in customer records.

Changes to Google Results APIs

SERP API provider SerpApi reported on 14 October that it had introduced an update intended to reduce the number of affected Google results returned through its service. The company also said it was continuing to investigate the problem.

A customer had previously requested a field indicating when a search returned no relevant results. The discussion highlighted that services using the API may otherwise need to inspect each response themselves.

AccuRanker’s API product, AccuSERP, returns the results page Google provides for an individual query and cannot perform the same historical comparison as the company’s rank tracker.

According to AccuRanker, AccuSERP shares browser sessions with its rank-tracking platform and prioritises sessions that have passed the tracker’s checks and continued to produce reliable results.

What SEO Teams Should Do About Ranking Drops

Before reporting major ranking declines from late September or October, SEO teams should establish how their tracking provider handled the affected period.

For example, SEOmonitor’s volatility safeguard may retain an earlier ranking for up to two daily updates while it checks whether a sharp decline continues. AccuRanker users can request a fresh keyword check, while other validation methods can help identify unusual changes.

Google Search Console can provide additional evidence when investigating ranking movements. However, its average position metric is not directly equivalent to a rank tracker’s position for a specific query. Google explains that search results can vary according to the device, location, time and search history involved. A high average position in Search Console therefore does not, by itself, prove that another platform has recorded an incorrect ranking.

If an agency needs to resend reports covering September or October, it should ask the provider whether historical records for those dates have been revised. Older exports may no longer match the figures currently shown in the platform, so both versions should be checked before being shared with clients.

Teams retrieving Google results through APIs should also review the safeguards available at their chosen endpoint. Depending on the service, these may include screened browser sessions, historical comparisons or internal checks to validate responses before they are used in reporting.

Questions Remain About the Extent of the Problem

As of 22 October, the providers discussed had not cited a public statement from Google explaining the issue. They have also described their own validation methods without providing independent testing of how effectively those methods identify inaccurate results.

The figures published by different companies are not directly comparable because they measure different things. SEOmonitor’s estimate of around 8%, for example, concerns the Google results it receives, while AccuRanker’s historical check examines a sample of stored keyword-days.

These distinctions matter when assessing how widespread the problem is and how much confidence to place in each estimate.

What Happens Next?

SEOmonitor says it will update its incident information as Google’s changes develop and its own improvements are introduced. Ahrefs has also said it will provide updates through its dashboard.

SEO teams should check for these updates before relying on rankings from potentially affected dates in client reports or management reviews. Confirming whether historical data has changed can help prevent outdated figures from being mistaken for current performance.