Google’s Spam Update vs. AI Sites: An SEO Experiment
▼ Summary
– The author conducted an experiment by creating three low-quality affiliate websites using AI-generated content, public data, and aggressive SEO tactics, which initially gained traffic.
– These experimental websites were quickly and completely de-ranked by a Google spam update, demonstrating that such low-trust, manipulative tactics are not sustainable.
– The core takeaway is that sustainable affiliate marketing still works, but only as a monetization layer for websites built on trust, expertise, and people-first content, not as a sole growth engine.
– The future of content, particularly by 2026, will favor deep, verticalized research and thought leadership that builds community discourse over easily replicated, ranking-focused pages.
– Successful content strategy will depend less on a single channel like search and more on creating unique value across multiple owned and partner channels to build a defensible business moat.
There was a time when ranking websites felt almost formulaic. Targeting specific commercial queries with exact-match domains and structured content could generate substantial affiliate revenue with surprisingly little ongoing effort. The focus was often on technical updates for freshness rather than deep value. This approach, while once viable, has fundamentally shifted. A recent personal experiment highlights this change starkly, contrasting a trusted, authoritative website with a fleet of purely AI-generated sites built to test the limits of current search algorithms.
The initial, successful model was built on a foundation of trust. For an established website in a “Your Money or Your Life” (YMYL) category, we scaled affiliate efforts responsibly. We hired subject matter experts to create genuinely helpful, educational content that targeted commercial keywords but served a broader informational purpose. This new content complemented thousands of pages of legacy user-generated content, creating a valuable ecosystem with brand authority, original research, and expert insights. It represented a sustainable, helpful experience for visitors.
The experiment aimed to strip away every element of that trust. Inspired by tactics promoted online, I purchased three partial-match domains targeting queries like “best welding schools.” The goal was to intentionally test low-trust, high-scale methods. Using AI, I generated thousands of bottom-funnel pages by templating public data, creating directory listings and targeting superlatives by program and state. The sites had zero brand signals, no original research, and employed aggressive internal linking for crawl coverage over user intent. They were pure search engine fodder.
Initially, the system responded. Indexation was rapid, and the sites began generating clicks for long-tail queries, amassing about 200 in-market clicks each within months. However, the first major Google spam update after their launch completely flatlined their traffic, dropping clicks to zero. Attempts to revive them with data updates failed. The lesson wasn’t merely that the sites failed; it was that Google’s systems tolerated them just long enough to learn from the patterns before deeming them unhelpful. Their only value was ranking, and once that signal vanished, nothing of substance remained.
This raises a critical question: does affiliate content marketing still function? The answer is nuanced. It works as a monetization layer for established, helpful sites, but not as a primary growth engine. Google’s guidelines are clear: content must be created for people first, not to manipulate rankings. Using automation like AI primarily for ranking manipulation violates spam policies. Even with best practices, the landscape has grown more challenging due to AI proliferation and other market shifts.
The core insight extends beyond spam updates or affiliate marketing. Businesses reliant on a single, easily replicable distribution channel are profoundly vulnerable when that channel evolves. The future of content is moving away from creating countless pages to rank for transactional queries. Instead, the focus is shifting toward verticalized research, original benchmarks, and insights that spark genuine discourse within specific communities. Content becomes a blend of discovery, discussion, and thought leadership that spans multiple channels.
Consider a fintech SaaS company. Instead of publishing pages targeting “best financial forecasting software,” they could conduct deep-dive research with industry leaders to identify major market gaps. Validating whether their product solves these problems creates a powerful wedge into the community. These insights fuel interactive assessments and benchmark reports that help organizations understand their position. While this content may not rank #1 for high-volume searches, it can be leveraged through owned channels, partnerships, and paid media to reach ideal clients. Companies like Stripe with its “Developer Coefficient” report exemplify this strategy.
This model demands a different mindset. It involves slower feedback, less directly attributable ROI, and fewer quick wins. It requires investment in distribution and partnerships. The future points toward fewer pages, deeper insights, a stronger point of view, and creating assets that are difficult for AI to replicate. The spam update exposed the razor-thin margin for anything built without trust. Ultimately, sustainable search marketing is less about avoiding penalties and more about building unique value that cannot be easily copied.
(Source: Search Engine Land)
