What the June 2025 Google core update actually punished (and what it rewarded)

What the June 2025 Google Core Update Actually Punished — And What It Rewarded

The rollout took seventeen days. By the time it finished, some sites had lost 60% of their organic traffic. Others — sites that hadn't changed anything in months — saw rankings they'd been chasing for years.

The Google June 2025 core update wasn't a tweak. It was a recalibration of what Google considers worth ranking. And the pattern of winners and losers tells a story that matters if you're publishing content in the second half of this year.

What Actually Got Hit

The sites that lost the most share three characteristics. Not one. All three.

First: content that could have been written about any company in the industry. Generic explainers using industry terminology without referencing specific products, services, or real situations. The kind of article that says "enterprise software solutions" instead of naming the actual product being discussed.

Second: publishing velocity that outpaced the site's actual expertise. Sites that went from 4 articles a month to 40 — with no corresponding increase in depth, original research, or business-specific detail. Google's systems got better at detecting when scale replaced substance.

Third: thin author attribution. Bylines with no verifiable expertise. No LinkedIn profiles. No evidence the named author has ever worked in the field they're supposedly explaining. E-E-A-T stopped being a checklist item and started being a ranking factor with teeth.

The combination mattered. Sites with generic content but strong author signals held steadier. Sites with specific content but no author presence took smaller hits. But sites with all three problems — generic, scaled, anonymous — those fell hard.

The Helpful Content System Finally Has Consequences

Google announced the helpful content system in 2022. For two years, it felt like a warning more than a weapon. Sites kept publishing AI-generated filler and ranking fine.

That changed. The June 2025 update integrated helpful content signals more deeply into the core ranking algorithm. The classifier that identifies "content created primarily for search engines rather than humans" now carries more weight.

What does that mean in practice? Articles structured around keywords rather than questions real people ask. Content that covers a topic without adding perspective, experience, or information unavailable elsewhere. The kind of blog post that exists because someone's SEO tool said the keyword had volume — not because the business had something worth saying about it.

The update didn't penalise AI-generated content as a category. It penalised content that reads like it was produced to fill a content calendar rather than help a reader. The tool matters less than the output.

What the Winners Did Differently

The sites that gained — and some gained significantly — share patterns worth studying.

They published less. Not because they were lazy, but because every article had a reason to exist beyond "we need to post something this week." Lower volume with higher specificity beat the reverse.

They referenced their actual business. Product names. Customer situations. Internal terminology. The content couldn't have been written by someone who'd never worked there, because it used information only someone inside would know.

They showed expertise through specificity, not claims. Instead of saying "we have years of experience in this industry," they demonstrated it by explaining edge cases, sharing real examples, and taking positions that generic content avoids.

They built topical authority deliberately. Not by covering everything tangentially related to their keywords, but by going deeper on fewer topics. A site about commercial HVAC that published 30 detailed articles on ventilation systems outranked competitors with 200 surface-level posts on every HVAC topic.

The AI Content Question

Every update brings the same question: is Google penalising AI content?

The answer remains the same, but the stakes are higher. Google isn't detecting whether AI wrote something. It's detecting whether the content is worth ranking. AI-generated content that sounds like AI-generated content — generic, safe, interchangeable — ranks worse because it's less useful. Not because of how it was made.

The distinction matters. A human can write generic filler. AI can produce specific, useful content if given the right context. That's exactly what BrandDraft AI was built for — it reads your actual website before generating anything, so the output references your real products and terminology instead of industry placeholders.

The June 2025 update didn't change the rules. It enforced them more strictly. Content that helps real people with real questions from a credible source still wins. Content that exists to capture search traffic without earning it loses more decisively than before.

What This Means for the Rest of 2025

If your traffic dropped, audit for the pattern: generic + scaled + anonymous. Fix the combination, not just one element.

If your traffic held steady, don't assume you're safe. The update rewarded specificity and punished its absence. Steady might mean you're on the border.

If your traffic increased, study what worked. The articles that climbed probably share characteristics: they reference your actual business, they demonstrate expertise through detail rather than claims, they answer questions thoroughly enough that readers don't need to click back.

The core update winners losers list will keep shifting as the dust settles. But the direction is clear. Google wants content written by people who know what they're talking about, for readers who actually need the information, about specific topics where the publisher has genuine authority.

Everything else is noise. And the June 2025 update turned the volume down on noise significantly.