What did Ahrefs actually measure?

Ryan Law, Director of Content Marketing at Ahrefs, wrote that each trend is "not speculation" because it "shows up in search interest itself." His team analyzed one million pages across one hundred thousand search results. Fully AI-written pages appeared at position one, with AI output populating the top three.

Heavier AI use lined up with slightly lower rankings. The authors attribute that gradient to content quality, not to algorithmic penalty against AI generation.

The study title states it bluntly: Google does not punish AI content, it punishes bad content. One caveat remains: Ahrefs itself sells Brand Radar, the AI visibility tool named in the article, and the underlying methodology does not appear in the public version for any reader to verify independently.

Are AI visibility tools a real category?

Search demand for AI visibility tools climbed 266% year over year to 2,100 monthly queries. Related terms follow the same pattern:

• "AI visibility tracker": up 156% to 2,400
• "AI visibility": up 145% to 4,400
• "LLM visibility": up 126% to 1,200

Most of this category did not exist before 2025.

Generative Engine Optimization reached 7,200 monthly queries, up 84% year over year. The vocabulary forms in real time across multiple related terms in this newly emerged segment.

The growth rate is steep, yet absolute volume remains modest against established SEO terms, and new categories routinely show exaggerated early growth before plateauing in the longer view.

How much click traffic has Google lost?

In a cohort of roughly 75,000 sites, Google's share of referral traffic fell from 35% to 24% in about a year. Ahrefs also reported that AI Overviews cut clicks to the top-ranking result by 58%.

ChatGPT accounts for 0.32% of referrals and carries essentially the entire AI category. Claude, Perplexity, and Gemini register as rounding errors in traffic terms.

The measurement gap matters most here, since AI tools take content from publishers, deliver answers to users, and return no visible click that marketers can actually trace through conventional dashboards anymore.

Is agentic commerce a real channel?

"Agentic AI" reached 111,000 monthly searches, up 37% year over year. "AI agents" hit 46,000. Agentic commerce climbed 16% to 6,400.

Marketing to agents specifically barely registers in search data. Terms like "marketing to AI agents" and "agent experience optimization" only appeared in 2025. Volumes are negligible today. Direction matters more than size at this early stage.

The opportunity appears genuinely unclaimed, yet whether agent referrals ever produce measurable traffic at scale is a question that Ahrefs does not actually address in any portion of its source data.

What role changes does the data support?

Job titles that did not exist in 2024 now register real search demand:

• "AI marketing director" — first seen 2025
• "Head of AI marketing" — first seen 2025
• "Marketing AI engineer" — first seen 2026

Volumes remain small. The vocabulary materializes out of nothing.

"Prompt engineer," the 2023 poster child for AI work, already folds into other titles as the specific label fades while the underlying skill of system design persists across functions.

What should marketers monitor next?

Track AI Overview click share quarterly, since Google's surface mix changes with every core update. Watch whether AI referral traffic concentrates in one engine or spreads across several.

Monitor whether synthetic creator content converts at parity with human-produced creative. The cost case is strong. The performance case remains unproven at scale.

The category is real, the measurement is incomplete, and the next twelve months will determine which tools and which definitions actually survive in the wider marketing stack across the industry.