Live · updated just now

Google ranking data glitches and a website speed tool update this week

Monday, 9 November 2026 · Marketing & design · Full report

What was found: A weekly industry roundup 1 reports three changes affecting how websites are measured and crawled. First, rank-tracking tools have become less reliable. Google has started returning results built around a single word from a search query. In an SE Ranking review, these accounted for 84% of flagged cases, and 96.5% of them appeared at position 9 or lower. Nozzle found that the share of affected results pages climbed to 7.6% by late October. Google has also made its "goto" redirect links harder to decode, and DataForSEO can now resolve only about half of them. Second, PageSpeed Insights moved to Lighthouse 13.6.0. This version changes how its simulated page loading treats render-blocking resources and fetch priority, so lab scores can shift, particularly for sites that rely heavily on JavaScript. Third, Google, Microsoft, and OpenAI said they plan to support a shared robots.txt rule, `__usage_ai-training`, that lets site owners state whether their content may be used to train AI 1.

Who it applies to: The findings matter for SEO professionals, website owners, and companies that build rank-tracking tools. They also matter for anyone who checks site speed in PageSpeed Insights or wants control over AI training on their content. The ranking data came from Google US desktop searches only 1.

How strong the evidence is: Moderate. The full article was read. It draws on industry analysis, data from several tools, status reports, and company announcements. SE Ranking reviewed 10,000 Google US desktop keywords, and the article adds data from Nozzle, DataForSEO, Serper, and SerpApi 1. Each tool measures the problem in its own way, so the reported impact varies from one to another. The robots.txt rule is still an Internet-Draft, and none of the three companies supports it yet. The article's text also ended mid-sentence, so part of it may be missing.

What it means in practice: If your rankings at position 9 or lower have dropped since late September, check whether the results now show unrelated single-word matches. Compare your tracker's figures with Google Search Console before drawing conclusions. When comparing PageSpeed lab scores before and after a change, make sure both reports used the same Lighthouse version, and give more weight to real-user field data than to lab scores. Site owners should also get ready to adopt the AI training rule once the companies actually put it into effect 1.