Google's third spam update of 2026 started rolling out August 18. Here's what SpamBrain is targeting, how it differs from the March and June updates, and the specific changes that move rankings.
Google's August 2026 spam update began rolling out on August 18. It's the third spam enforcement action of 2026, following updates in March and June, and it's the first to explicitly target scaled AI content at the page level rather than the domain level.
If your rankings moved in the last week of August, this is the most likely cause. Here's what's actually happening and what changes actually matter.
The March, June, and August updates each targeted different abuse patterns. Treating them as the same "spam crackdown" is why most advice on this topic misses the point.
March 2026 (fast — completed in ~19.5 hours): Targeted manipulative redirects and cloaking at scale. If a page showed Googlebot different content than human visitors, March caught it. This update was surgical and fast because the signal was unambiguous.
June 2026 (slower, 3-day rollout): Focused on scaled low-value content and keyword stuffing at the page level. The June update is what initially pressured a lot of programmatic content sites. It's also when Google clarified publicly that the spam system and core quality system are separate — a point worth understanding, because the remedies are different.
August 2026 (ongoing): Hits every language and region simultaneously. Early data from SEMrush and Sistrix shows volatility concentrated in health, finance, and legal verticals — the classic YMYL categories where Google's quality bar is highest.
The pattern across all three: Google is running SpamBrain on more granular signals than it did in 2024 and 2025. The days of site-wide penalties being the primary mechanism are largely over. It's now page-level scoring.
SpamBrain is Google's AI-based spam detection system. It classifies content rather than just checking for rules. That distinction matters because it means there's no single flag to avoid — it's a holistic assessment.
The signals it weights most heavily in 2026, based on what's recovering versus what isn't across the SEO community:
Original information contribution. Google is now explicitly evaluating how much genuinely new information your page adds compared to what already ranks for the same query. Rephrasing existing top results, even well, doesn't satisfy this. The bar is: does this page contain data, analysis, or perspective that a searcher could not get from the existing top 3?
Scaled content from templates. Pages that follow an identical structure across hundreds of URLs — especially if that structure maps to a keyword template — are at much higher risk in August than they were before. This applies whether the content is AI-generated or human-written to a formula.
Hidden text and cloaking (still): These carry over from March. If you had sites caught in March that didn't fully fix these issues, August is likely compounding the damage.
Thin interstitials and doorway pages. Pages whose primary purpose is to rank for a keyword and funnel traffic elsewhere, without providing standalone value, are a specific focus in the August update.
What SpamBrain is not doing in August:
This is the most practically important thing to understand. Google said it directly after the June update: the spam system and the core update quality system are separate, with separate remedies.
If you were hit by the spam update, the issue is a named policy violation — cloaking, scaled content abuse, manipulative behavior. Recovery requires removing or correcting the specific violation. Improving content quality won't help if the issue is a technical abuse pattern.
If you were hit by a core update, the issue is that Google reassessed the relative quality of your content against competitors. Recovery means producing genuinely better content — more expertise, more original data, more depth. Fixing technical issues won't help here.
The practical test: look at your Search Console data. Did rankings drop gradually around the core update dates (March core update, May core update), or did they drop sharply on August 18-19? A sharp drop that correlates exactly with August 18 points to spam. A gradual drift that started earlier points to core quality.
Short answer: fixing the specific violation, not general SEO improvements.
For scaled low-value content: Don't just improve a sample of pages. Google's systems evaluate the pattern across your entire site. If you have 800 pages that follow a keyword-stuffed template and you improve 50 of them, the site-level signal barely moves. You need to either consolidate those pages into fewer, genuinely useful resources, or remove them entirely and redirect to something that earns its place in the index.
For cloaking or hidden text: Fix it and submit the affected URLs for recrawl. This is mechanical — once the violation is gone, recovery can happen within weeks.
For doorway pages: The hardest category. If pages exist primarily to rank rather than to serve a user need, the fix is to either rethink what the page does for a visitor or redirect and consolidate.
For keyword stuffing: Read the page out loud. Any phrase that sounds weird to a human is stuffed. Remove it. This sounds obvious but a lot of keyword stuffing is invisible when you're editing it; reading aloud catches it.
Based on patterns from the March and June updates:
One thing worth noting: Google's spam team has a manual review process for sites that believe they were incorrectly classified. If your site doesn't fit the violation categories above and you saw a significant drop on August 18, it's worth filing a reconsideration request with specific evidence that your content practices don't match the violation definitions.
Three spam updates in one calendar year is not normal. 2024 had two. 2023 had three. The frequency in 2026 suggests Google is running SpamBrain on tighter enforcement cycles — possibly responding to the volume of low-quality content that AI tools made economically viable to produce at scale.
If you weren't affected in August, that doesn't mean your content is invulnerable. The pattern suggests another update before year-end. The best preparation is a content audit focused on pages that exist primarily for rankings rather than for visitors: thin category pages, auto-generated location pages, glossary entries built from keyword research rather than genuine expertise.
Cut or consolidate the ones that don't justify their existence. The sites that are weathering 2026 cleanly aren't doing anything special — they're just not building content whose primary audience is a crawler.
MarketerTools
Marketing Practitioners
Written by the MarketerTools team — practitioners who build and use tools for marketers every day.
Put this into practice
Use the free tools on MarketerTools to apply what you just read.
Browse all tools →