Tech publications lost 58% of their Google search traffic between 2024 and 2026, according to new data from Growtika. This isn't a slight downturn. It's an existential collapse.
The culprit isn't a mystery. Google's AI Overviews now answer most tech questions directly in search results, eliminating the need to click through to actual articles. Why read a review of the new iPhone when Google can summarize twelve reviews for you in a neat little box?
The technology is impressive. The question is whether journalism can survive it.
I've been in the startup trenches. I understand disruption. But this isn't creative destruction—it's something more insidious. Google is using content created by publishers to train AI models that then compete with those same publishers for traffic. It's the search engine equivalent of learning to cook from your recipes, then opening a restaurant next door and giving away your dishes for free.
The data from Growtika shows the damage isn't evenly distributed. Major outlets like The Verge, TechCrunch, and Ars Technica have seen traffic crater, but they have diversification strategies—newsletters, events, membership programs. Mid-tier publications that depended almost entirely on search traffic are quietly shuttering.
Here's what makes this particularly brutal: tech publishers were early adopters of SEO best practices. They invested heavily in structured data, fast-loading pages, and the kind of technical excellence Google claimed to reward. They played by the rules. And now those same optimizations are making it easier for Google to extract and repackage their content without sending traffic back.
Some publications are fighting back. A coalition of tech media companies is exploring legal action, arguing that AI Overviews constitute copyright infringement. Others are de-indexing their content from Google entirely, betting that direct traffic and social media can compensate. (Spoiler: it mostly can't.)
The more interesting response is happening at the business model level. Publications are realizing that if is going to summarize their content anyway, they need to produce things that be summarized. Deep investigations. Original reporting. Analysis that requires understanding the industry, not just pattern-matching on training data.

