A fundamental shift in how email is consumed began in 2026 that email practitioners have not yet fully reckoned with: AI systems at Gmail and Apple Mail now read email before humans do, summarise its contents, and in many cases provide the recipient with enough information to satisfy their curiosity without ever opening the email or clicking any link. This is the zero-click email phenomenon — email that is technically delivered, technically opened (by the AI), and technically read (by the AI's summarisation engine), but that generates no human click because the AI summary provided everything the recipient needed to know. The commercial implications are significant, the measurement implications are acute, and the strategic adaptation is urgent for any programme where click-through generates revenue.
The Zero-Click Email Phenomenon in 2026
The zero-click email phenomenon emerged in search first — Google's AI Overviews in search results allow users to get answers to their questions without clicking any search result. The same dynamic has now arrived in email: Gmail's Gemini AI Inbox summarises email content in the inbox preview, allowing recipients to understand an email's content without opening it or clicking any link. When a recipient reads the Gemini-generated one-line summary of a newsletter article and understands the key point, they may not click through to the full article on the newsletter's website — the summary delivered sufficient value.
This is not a hypothetical risk — it is a measured, observed phenomenon. Folderly's analysis following the Gmail Gemini rollout documented a 10% decline in click-through rates across their monitored email programmes (from a 4.35% average to 3.93%). This decline occurred with stable delivery rates and stable inbox placement rates, pointing directly at AI-mediated consumption — not deliverability degradation — as the causal factor. The clicks are declining not because email is being blocked, but because AI summaries are capturing the informational value that previously required a click to access.
The zero-click phenomenon compounds the already-established trend of AI-generated opens from Apple MPP and Gemini AI processing. In 2026, a recipient's complete email experience may consist of: (1) AI automatically processes the email at delivery — generating an open event and a summary. (2) Recipient glances at the AI-generated summary in the inbox list view. (3) Recipient does not open the email — the summary was sufficient. (4) Recipient does not click any link — no website visit, no conversion opportunity. The sender's analytics shows: Delivered ✓, Opened ✓ (machine-generated), Clicked ✗, Revenue: $0. The programme appears to be performing well on delivery and open metrics while quietly losing the clicks and conversions that generate commercial value.
The CTR Data: 4.35% to 3.93% — What It Means
The 10% CTR decline that Folderly documented post-Gemini rollout is the tip of a trend, not the full extent of it. The Gemini rollout in February 2026 affected the fraction of Gmail users who had access to the AI Inbox feature at launch — which was not 100% of Gmail users immediately, but a growing rollout. As Gemini AI Inbox reaches full Gmail penetration (expected through 2026), the CTR impact will grow proportionally with adoption.
The 10% decline is also averaged across all content types and programme types. Content that is fully summarisable by AI (informational newsletter articles, product announcements, market updates) has experienced higher CTR decline than content that is not summarisable (personalised data, exclusive access links, interactive content). Programmes that are disproportionately content-heavy (newsletters, media publishers, content marketing email) have experienced more severe CTR decline than programmes that are primarily transactional or offer-based (e-commerce promotions, SaaS trial emails, appointment confirmations).
What a 10% industry-average CTR decline means for individual programmes depends on the programme's content type, audience AI adoption rate, and current CTR baseline: for a newsletter programme at 2% CTR, a 10% decline means moving from 2.0% to 1.8% — losing 2 clicks per 1,000 sent. At 1,000,000 sends, this is 2,000 fewer clicks per campaign, which may represent $200-2,000 in lost advertising revenue or conversion revenue depending on the programme's monetisation model. The aggregate commercial impact, compounded across weekly sends over a year, is material for programmes dependent on click-driven revenue.
Which Email Content Drives Zero-Click Consumption
The zero-click phenomenon is not uniform across all email content types. Content that is most susceptible to AI summary-driven zero-click consumption:
Self-contained news summaries: Newsletter articles that completely summarise a news story in the email body allow the AI to create a perfect summary that delivers the full informational value — leaving no "incomplete picture" that requires a click to complete. If the article's key insight, conclusion, and recommendation are all in the email body, the AI summary captures them accurately, and the recipient has no information-seeking reason to click through to the website.
Market commentary and analysis: Investment and financial commentary that presents a clear thesis with supporting arguments and a conclusion is highly summarisable. The AI can accurately represent "Tech stocks are overvalued due to X, Y, Z factors. Recommendation: reduce exposure." without the recipient needing to read the full analysis article. The informational value is captured in the summary.
Product announcements with complete feature descriptions: "Introducing Feature X — it does Y and Z, available now at [price]." Fully summarisable. If the announcement email includes all relevant information (what it is, what it does, what it costs), the AI summary captures it and the recipient may not click to learn more.
Content that is most resistant to AI zero-click summarisation: (1) Exclusive content that exists only behind the click — the email says "I've published a deep analysis of X" but the analysis itself is only on the website/members page. The AI summary says "Article about X available at link" — which is not enough information to satisfy the recipient's curiosity. (2) Personalised data the AI cannot access — "Your specific usage data for March is ready" but the data is only in the account dashboard. The AI cannot generate the data summary because it doesn't have access to the data. (3) Time-sensitive decisions — offers with actual deadlines, event registrations with capacity limits, decisions that require immediate action. The AI summary communicates the urgency, which motivates clicking rather than satisfying curiosity without the click. (4) Interactive content — polls, surveys, quizzes — that the AI can describe but cannot fulfil.
Newsletter Publisher Revenue Impact
Newsletter publishers whose revenue model depends on website visits — advertising revenue from ad impressions, subscription conversion from website landing pages, affiliate revenue from click-through purchases — face disproportionate commercial exposure from the zero-click AI summary phenomenon. When Gemini AI summarises a newsletter article and the subscriber reads the summary without clicking through to the publisher's website, the publisher receives: zero advertising impressions (no website visit, no ads served), zero subscription conversion opportunity (the subscriber satisfied their need without reaching the paywall), and zero affiliate revenue (no click-through to the product being recommended).
The newsletter publisher adaptation to zero-click AI: the core insight is that full-article-in-email design — which provides maximum reader value in an email-only world — is the highest-vulnerability design in an AI-summary world. The newsletter article that is completely readable in the email body requires no click-through and generates zero website revenue. The newsletter article that is previewed in the email (3-4 paragraphs of context and setup) but requires a click to read the conclusion, analysis, or recommendation generates a motivated click from the reader who wants the complete information.
This is the "truncated article" approach that many publishers adopted when they first began gating their content behind website visits. The zero-click AI era creates a new reason to employ this approach: the AI can summarise the preview section, but it cannot summarise the full article because the full article is not in the email. The AI summary says "This article discusses X and promises analysis of Y — read more at [link]." The subscriber who wants Y has to click.
Content That Earns Clicks Despite AI Summarisation
The click-resistant content strategy builds irreducible engagement into every email — content that requires human action to access and cannot be satisfied by an AI summary alone:
Exclusive, gated content: The most reliable click-resistant email type is one that explicitly promises content available only at the click destination. "I spent 8 hours analysing competitor pricing strategies — my full findings are only in the linked report." No amount of AI summarisation can deliver the 8-hour analysis. The click is required. The email teases the analysis without delivering it.
Personalised account data: Anything tied to the recipient's specific data — usage statistics, performance benchmarks compared to peers, personalised recommendations based on their history — cannot be fully represented by an AI summary because the AI does not have the underlying data. "Your email programme delivered 14% below your industry's top quartile this month — here's your personalised improvement plan." The specific data and the specific plan require the click.
Community and human elements: Reader responses, Q&A, live events, and community discussions that involve real-time human participation cannot be summarised into passive consumption. A newsletter that includes "This week's most-voted reader question was X — 47 readers shared their take, here are the top responses" creates social engagement motivation that an AI summary cannot satisfy. The summary says "There are reader responses to question X" — the human wants to see the actual responses.
Time-sensitivity with genuine stakes: Early-bird pricing that expires, webinar seats that fill up, one-day offers — these motivate clicks even when the AI summary accurately communicates the deadline. The AI summary says "Sale ends tomorrow" — this increases click motivation rather than satisfying the need without the click.
The Deliverability Engagement Paradox
The zero-click AI phenomenon creates a specific deliverability paradox: the AI systems that are causing click rates to decline are the same AI systems that evaluate email quality for inbox ranking. Gemini AI ranks inbox visibility based on engagement signals — and AI-mediated zero-click consumption reduces the human engagement signals that determine future inbox ranking. A subscriber whose experience of an email is entirely AI-mediated (the AI reads it, generates a summary, the human sees the summary without opening or clicking) contributes nothing to the positive engagement signals that maintain High Gmail domain reputation.
This means the zero-click phenomenon is self-reinforcing from a deliverability perspective: AI summarisation reduces clicks, lower clicks reduce engagement signals, lower engagement signals reduce AI inbox prioritisation (Gemini deprioritises the sender for future sends), lower inbox prioritisation further reduces click opportunities, which further reduces engagement signals. The programme caught in this cycle needs to interrupt it by generating genuine human engagement — content that earns clicks despite AI summarisation — rather than relying on volume alone to maintain reputation signals.
The deliverability implication of zero-click AI is therefore not just about the business model of clicks — it is about the engagement signal economy that determines future inbox placement. Every campaign that generates only AI-mediated zero-click consumption is eroding the engagement signal foundation that keeps the programme in High Gmail reputation. Building click-resistant content that earns human engagement is simultaneously a revenue strategy (for programmes dependent on click-driven revenue) and a deliverability strategy (for all programmes dependent on engagement signals for inbox placement).
New Metrics for the Zero-Click Era
Standard email engagement metrics — open rate, click rate, click-to-open rate — were designed to measure a world where email engagement was binary: humans read and clicked, or they didn't. The AI-mediated inbox of 2026 requires metrics that can distinguish human engagement from AI-mediated consumption:
Human click rate: Clicks per delivered email, excluding machine-generated activity. Since link clicks are not currently inflated by AI systems the way open events are, click rate per delivered (CTD) remains the most reliable human engagement indicator. Monitor CTD, not CTO (which has an AI-inflated denominator).
Email client-segmented CTR: Track click rate separately by email client — Gmail, Apple Mail, Microsoft — to identify whether CTR decline is concentrated in specific AI-implementing clients. A CTR decline concentrated in Gmail while other clients are stable specifically implicates Gemini AI zero-click summarisation.
Revenue per delivered: The ultimate commercial performance metric. CTR measures engagement; revenue per delivered measures commercial outcome. If CTR is declining but revenue per delivered is stable, the lower-clicking recipients are not commercially valuable anyway. If both are declining together, the AI zero-click effect is reducing commercial performance, not just measuring changes.
Reply rate (for programmes that solicit replies): Unambiguously human — AI systems do not reply to email without explicit user intent. Monitor reply rate as a human engagement quality signal even at the low rates (0.1-0.5%) that most commercial programmes achieve.
The Strategic Adaptation: Email as Distribution, Not Destination
The most resilient strategic posture for the zero-click AI era reframes what email is for: email is a distribution channel for driving human attention to value that exists elsewhere — on the website, in the member portal, in the community, in the product — rather than a destination where the full value is delivered within the email body itself. This framing change has profound implications for email content design.
Under the old model: the newsletter delivers the full article. The email is the destination. Value delivery is complete within the email. Under the new model: the newsletter distributes the promise of the full article. The email is the distribution channel. Value delivery requires the click. The AI can summarise the promise; it cannot summarise the value that exists only at the click destination.
The practical transition: identify which elements of current email programmes can be moved to click-gated destinations (full analysis, detailed data, community interaction, premium content) versus which must remain in the email body (enough context to motivate the click). The email body becomes the teaser, the hook, the context — not the complete package. The complete package is one click away, accessible only to humans who chose to engage, and invisible to AI systems that can only read what is in the email.
The zero-click world is a world where email's role in the content and commerce value chain becomes more clearly defined: email is the distribution mechanism that moves attention, not the destination that delivers value. Programmes that understand this shift and design for it — creating email that distributes compelling promises of value available at the click destination — will maintain engagement in the AI-mediated inbox. Programmes that continue designing email as a self-contained value delivery mechanism will see their click rates continue declining as AI systems complete more of the consumption task before the human ever has to engage.