Google Search referrals to the web have plummeted, AI links are ‘less than 1%’ of traffic

Google Search referrals to publishers dropped 34% year-over-year (small publishers -60%, medium -47%, large -22%), Google Discover fell 15%, and overall publisher traffic declined 6% between 2024 and 2025. AI chatbots account for <1% of publisher referral pageviews despite ChatGPT referrals rising >200% in 2025; some tech media sites saw extreme declines (Digital Trends -97%). Chartbeat/Axios data conflicts with Google’s claim of stable organic click volume, signaling sustained downside risk to digital publishers' traffic and monetization that favors dominant platforms and may pressure smaller publishers' revenues.

Analysis

AI intermediaries are eating the discovery layer: the dominant change is not a one-off traffic reallocation but a durable compression of organic click-through that disproportionately removes tail queries and low-friction pageviews. Over 6–24 months this will shrink the marginal supply of fresh, niche content (the long tail) that underwrites programmatic ad inventory, raising CPM volatility as buyers concentrate spend on fewer premium placements and on-platform surfaces.

Winners will be owners of the AI layer and closed ecosystems that can monetize impressions without passing through to third‑party sites; losers are the mid/small publishers and the programmatic exchanges that relied on broad, cheap supply. Second‑order exposure includes CMS and hosting vendors that depend on publisher scale (potential downside to cloud usage per publisher) and ad tech firms whose pricing models assume large, low-CPC inventory — those cost structures will be renegotiated or fail.

Reversal catalysts are concrete and measurable: regulatory mandates forcing click-through to verified sources, publisher licensing deals that embed paid attribution, or rapid UX fixes in AI products that materially raise referral CTRs. Each catalyst has different timing: regulatory/legal outcomes could take 12–36 months, platform UX/product fixes can appear in quarters, and publisher consolidation effects will play out over 1–2 years.

From an investment lens the market likely underprices the fragility of Google’s ad “quality” narrative given reinvestment and margin sensitivity; conversely, social/community platforms are being underfollowed as potential aggregators of fact-checked referral traffic. Watch three metrics weekly/monthly: organic click-through rate from AI surfaces, publisher CPMs and fill rates, and Google’s search ad revenue per query — these will be the leading indicators for re-rating risk.

Key Decisions for Investors

• Short GOOGL via a 6–12 month put spread (sell 1x 6–12 month OTM puts, buy 1x deeper OTM puts) sized to 1–2% portfolio risk; target 12–18% downside if marketplace CPMs compress and ad revenue growth decelerates. Rationale: asymmetric downside with capped cost if AI continues to internalize queries; stop-loss if Google reports sequential ad CPM recovery or guides sustainable click monetization upside.
• Pair trade — short GOOGL equity (or increase the put-spread) and go long RDDT equity (size smaller, 0.5–1% portfolio) over 9–18 months: RDDT is a thematic beneficiary if community-led verification/links become the default ‘last mile’ for AI fact‑checks, enabling re-monetization. Risk/reward: expect >1.5x upside on RDDT vs downside protection from the put spread; cut position if RDDT user engagement metrics stall for two consecutive quarters.
• Tactical options play: buy RDDT 12–18 month calls (LEAPS) at moderate strike to capture re‑rating on improved monetization or M&A interest from larger platforms; limit allocation to 0.5% portfolio as a high-volatility payoff. Exit or roll if monthly active user growth decelerates below consensus or if RDDT misses sequential revenue milestones.
• Watchlist triggers (no trade until met): a) Google makes measurable UI changes that materially increase referral CTR (>50% improvement on sampled queries) or b) a publisher consortium announces a licensing deal that captures >5% of search-originated ad budgets — either would materially reduce short conviction and warrant tightening stops.