Google closes September 2026 spam update as AI Overviews reach 83% of branded searches

Spam update lands after a 13-to-14-day rollout

Google officially closed its September 2026 spam update on October 8, 2026, ending what the company had flagged as a rollout that could take up to two weeks. One outlet recorded the exact completion time as 4:37 a.m. ET, with at least one independent tracker measuring the rollout duration at 13 days and 16 hours. A maintenance-grade time window matters here because sites do not only watch for ranking changes on day one — practitioners look for the several-day tail in which residual actions and re-inclusion requests resolve. With the update finished, every site owner whose traffic dipped during the window now has a confirmed boundary for before-and-after comparisons in Search Console.

UGC applications reopen as cleanup pressure lifts

Google opened user-generated content (UGC) applications again once the spam update wrapped, ending a pause that practitioners had tied to risk control during the rollout itself. The reopening changes workflow immediately: publishers who held review queues for contributor posts during September can resume publishing and re-indexation work, while those who kept UGC off the site now have a clearer window to test site-reputation shifts against the post-update baseline. Any team considering a return to UGC should audit contributor pages against the same spam signals the just-completed update targeted, since the same enforcement posture remains in place even after the rollout completes.

AI Overviews hit 83% of branded searches as GPT-6 ships interactive answers

AI Overviews were reported appearing on 83% of branded searches, a penetration figure that makes brand-question ownership a primary GEO concern rather than an experimental one. When AI answers surface for the majority of queries that name a brand, the text Google (or any answer engine) picks as its source is the page that owns the conversation — not the page that ranks first by traditional blue-link standards. Separately, GPT-6 was reported to have added interactive answers inside its conversational surface, raising the same brand-question concern for OpenAI's ecosystem.

Practitioners who have not yet built a brand-question baseline are now working against an answer layer that has already gone mainstream. Practical first reads on this: work through Generate GEO Brand Awareness Questions the Right Way to map the brand-question set, then Document Steps for GEO Brand Awareness Question Generation so the workflow survives hand-offs, Anchor Awareness Work With a GEO Brand Question Baseline for a repeatable answer frame, and How to Check GEO Brand Question Results After Generation for post-generation QA — and run the list through the GEO Brand Question Generator to keep the set consistent across runs.

Rank-tracker anomalies: 11% of daily checks returned unexpected SERPs

A rank-tracking vendor reported that Google returned unexpected SERPs on 11% of daily rank checks, a figure that lands directly on top of the spam update window. When one in nine SERP samples is materially different from expectation, two practical reactions follow: first, treatment of any single-day rank-tracker reading as authoritative should be suspended until samples cluster back to baseline; second, the same anomaly data is itself a signal worth exporting before normalising, because it captures the shape of the rollout rather than just the steady state. Practitioners who do not already keep an isolated "rollout-period" export should start one for the next confirmed core or spam event.

Where this leaves the practitioner

The update is closed, UGC is open, and the AI answer layer now sits at 83% of branded searches across at least one reported measure, so the only durable move available to a site owner today is to verify (a) the Search Console window for their property against October 8, 2026, (b) the state of any held UGC queue, and (c) the brand-question coverage across both Google AI Overviews and GPT-6. For teams that have not yet instrumented AI surfaces separately, the higher-value next step is to build a recurring brand-question audit rather than to chase reaction patches.

Sustained GEO work runs through Troubleshoot GEO Brand Question Generation From Awareness and Simplify GEO Brand Question Generation From Awareness Work, with the GEO Brand Question Generator handling the recurring question set.