220 Sites Scaled AI Content. Forty Are Already Bleeding Up to 95% of Their Traffic. The Pattern Is Familiar.
Lily Ray tracked 220 domains scaling AI content. At least 40 lost between 40 and 95% of organic traffic from January 2026 onwards. The boom-bust cycle is back, faster than ever.
Google has seen this movie before. Doorway pages, scraped review content, link farms. The script changes. The ending does not.
Lily Ray, the most-read SEO analyst working today, has published an analysis of 220-plus domains that scaled AI content production over the past 18 months. At least 40 of them saw organic traffic drop between 40 and 95% starting around January 20, 2026. Google did not name a core update. The crash happened anyway.
The pattern Ray identified is specific. The affected sites had published dozens, hundreds or thousands of self-promotional listicles. Each listicle named the publishing company as the number one best option in its category. One site had published 2,000 articles all of which crowned the same company first.
That is not a content strategy. That is a black-hat tactic with an AI veneer.
Lily Ray identified at least 40 sites with traffic drops between 40 and 95% between January and April 2026 after AI-content scaling
The deeper issue is that AI tools have collapsed the cost of producing this kind of content. What used to take a content farm of 100 writers two years now takes a single operator and an API key two weeks. The volume of low-quality content hitting the web is unprecedented. Google's algorithms are responding accordingly.
Why it matters
The signal to Australian marketers is clear. Volume strategies that worked even six months ago are now active risk. Two trends compound.
First, Google's spam policies were updated in May 2026 to explicitly cover AI search manipulation. The window for tactical content scaling closed before most agencies finished scaling theirs.
Second, the AI search platforms penalise the same patterns Google does. A page Google deprioritises also tends to be cited less by ChatGPT, Perplexity and Claude. The double-platform exposure makes the downside worse than a Google-only era penalty.
For brands working with content agencies on a 'publish X articles a month' KPI, the next review cycle is going to be uncomfortable. Output KPIs without quality controls were always a poor proxy. They are now actively dangerous.
What to do about it
Audit your published content from the past 12 months. Flag any pieces that explicitly compare you favourably to named competitors without third-party validation. These are highest-risk.
Identify articles that exist purely to rank for '[your brand] vs [competitor]' queries. Decide whether they add value or read as self-serving.
Move agency KPIs from publish-volume to traffic-delivered or rank-position-secured. Volume KPIs incentivise the wrong behaviours.
Build an expert citation layer into your content workflow. Real quotes from named people, not LLM-generated authority signals.
If your traffic dropped between January and April this year, do not wait for Google to recover it. Audit, remove, consolidate and republish. The longer the low-quality pages sit, the more they drag your domain authority down.
The boom-bust cycle in content marketing has a new accelerant. The bust phase is here.