Roughly 30-40% of your marketing influence happens in places you can't directly measure. AI Overview compresses informational click-through. Dark social moves through DMs and private groups. How to operate when measurement is structurally incomplete.
Sources & attribution. The zero-click and dark social discipline draws heavily on Rand Fishkin's SparkToro research, Similarweb data on AI Overview impact, and the marketing measurement work of Common Thread Collective and Group M.
The invisible majority of traffic
Zero-click search — Google answering the query directly in the SERP without sending a click — has been growing for a decade. AI Overview accelerated it. As of 2026, roughly 31% of US SERPs include AI Overview, and the click-through rate on traditional organic results drops 18-28% when AI Overview appears (multiple studies; the exact number varies by query type).
Dark social — the traffic and conversions that arrive via DMs, private Slack groups, podcast mentions, and copy-paste links — is also growing. Most analytics systems classify dark social as "direct" or "(none)" traffic, masking what's actually driving it.
Why this matters strategically
When 30-40% of your influence happens in places you can't directly measure, optimization-led marketing breaks. If you only invest in what you can attribute, you under-invest in the channels that drive the most pipeline. The teams winning in 2026 are the ones that:
Acknowledge the measurement gap explicitly.
Invest in brand and content that propagates in dark channels.
Use self-reported attribution (HDYHAU surveys) to triangulate.
Run incrementality tests instead of relying on last-click.
Measuring dark social
Three techniques work:
HDYHAU surveys. "How did you hear about us?" on signup and conversion forms. The most-accurate-imperfect measurement available.
Geo holdout tests. Run a campaign in some metros, hold others as controls. Compare pipeline.
Media Mix Modeling (MMM). Model channel contribution from aggregate data. Captures dark-social impact better than last-click.
Optimizing for zero-click
Some queries genuinely don't need a click — the user got their answer. Trying to fight this is wasted effort. The strategic response:
Optimize commercial-intent queries. AI Overview compresses informational queries hardest. Commercial queries ("X vs Y", "best X for Y") remain heavily clicked.
Earn brand mentions in AI answers. When AI Overview appears, study which sources it cites. Optimize content to be the cited source.
Build for downstream conversion. Brand-touched users from zero-click impressions still convert eventually. Treat AI Overview impressions as upper-funnel brand investment.
Optimizing for dark social
You can't optimize what you can't measure directly. But you can invest in what gets shared in DMs:
Highly-shareable formats. Quick-take posts, short videos, podcast clips, single-frame insights.
Brand mentions in podcast appearances. A podcast mention by a credible host drives weeks of dark-social discovery.
Quotable language. Specific, declarative statements get screenshotted and shared. Mush gets forgotten.
Easy-to-copy URLs. Short, clean, descriptive URLs get shared more than long parameterized ones.
Sources & further reading
SparkToro studies on zero-click search and dark social — Rand Fishkin's research.
Similarweb data on zero-click and AI Overview impact.
Group M and major holding company 2025 attribution reports.
Ahrefs research on AI Overview click-through rate impact.
Common Thread Collective and MarketerHire on HDYHAU implementation.
RGM operator notes on dark-social measurement 2024-2026.