Marketing when nobody clicks: the zero-click playbook

The audience moved first. ChatGPT reached roughly 900 million weekly active users in early 2026, about double a year earlier, and Google's Gemini app passed 750 million monthly users. McKinsey found in October 2025 that 50 percent of consumers already use AI-powered search intentionally as a primary way to find information and make buying decisions, and Aeolyft's 2026 U.S. Search Trends Report puts weekly AI use at 58 percent of Americans. One 2026 estimate counts about 45 billion AI assistant sessions worldwide, equal to around 56 percent of global search engine volume; treat that as an estimate, but the direction is unambiguous.

The interfaces moved with them. Google's AI Overviews trigger on roughly 48 percent of tracked queries by early 2026 according to third-party tracking, up about 58 percent year over year, with the honest caveat that panels disagree and some report 13 to 25 percent depending on the query set. ChatGPT Search processes an estimated 250 to 500 million weekly queries. Every one of those interactions produces an answer, and only a minority produce a click.

For B2B, the shift is sharper still. G2 found 51 percent of software buyers now start research with an AI chatbot more often than with Google, up from 36 percent seven months earlier. Forrester's 2026 Buyers' Journey Survey of 18,000 buyers found 94 percent used AI during their most recent purchase and 47 percent built internal business cases before ever contacting a vendor. The question is no longer whether your buyers use AI answers. It is what those answers say while you are not in the room.

The classic funnel assumed the persuasion happened on your property. Search produced a click, the click produced a session, and the session was where positioning, proof and pricing did their work. Zero-click environments relocate the persuasion into the answer. The assistant compares the options, states the tradeoffs, and frequently names a recommendation, all before any website is involved. Your landing page cannot win an argument it was never invited to.

The evidence that decisions conclude inside answers is behavioral. G2 found 69 percent of B2B buyers chose a different vendor than they originally expected because of AI chatbot output, and 33 percent bought from a brand they had never heard of before an AI named it. A third of purchases going to previously unknown brands is what it looks like when shortlists are assembled by an assistant rather than by a browsing session.

This is uncomfortable for teams whose entire reporting stack counts visits, and it is worth being precise about what changed: the influence still exists, and it simply stopped emitting the telemetry you were counting. A buyer who saw you recommended in three separate AI answers and then typed your URL directly looks like a direct visitor with no history. The primer on this new surface, and why it deserves a line in your reporting, is what is AI visibility.

Zero-click does not mean zero traffic; it means the surviving traffic is pre-persuaded. WebFX, analyzing 2.3 billion sessions across 2024 and 2025, measured AI referral traffic growing 796 percent, with AI-referred visitors converting about 1.2 times organic. Adobe found AI traffic converting 42 percent better in March 2026, and Semrush reported AI-driven visitors converting at roughly 4.4 times standard organic. Ahrefs found up to 23 times, a figure widely treated as an outlier and cited here with that caveat. The studies disagree on magnitude and agree on direction, and the honest read of all of them is in what the AI conversion studies really say.

Keep the volume honest too: AI referrals are still around 0.18 percent of all sessions in some panels, and much AI-sourced traffic hides in direct. But the pipe is growing fast, and unevenly: ChatGPT commands about 92 percent of trackable LLM referral traffic, and SE Ranking's study of 101,574 websites found ChatGPT referrals jumping 36.7 percent in May 2026 alone to an all-time high. Small, hot and compounding is a profile worth building for.

Inside Google, the same both-sides pattern holds. Organic click-through drops about 61 percent when an AI Overview is present, according to 2026 tracking studies, yet brands cited inside the Overview see around 35 percent higher click-through. The answer box taxes everyone standing outside it and pays the brands standing inside it, a dynamic unpacked in AI Overviews and your clicks.

The first play is earning mentions: being one of the names an assistant offers when a buyer asks an unbranded category question. Start by finding out where you stand, and be strict about the method: unbranded questions only, repeated runs, per-engine scores, because questions containing your brand name inflate the result by construction and single runs are noise.

Then work the sources the engines actually listen to. 5W Research found Wikipedia and Reddit together account for over a quarter of ChatGPT's U.S. citations, and Reddit's citation share grew roughly 73 percent in commercial categories across 2025 and 2026. Independent reviews, comparison articles, community threads and reference entries teach assistants that your brand belongs to your problem space. Authentic community participation compounds here, and the rules for doing it without getting burned are in a Reddit strategy for brands that AI engines respect.

Mentions are the currency that pays inside clickless journeys, which makes them the primary zero-click asset. Their mechanics, and how they differ from citations, are laid out in mentions vs citations. The short version: mentions are earned mostly in other people's text, they build slowly, and they persist once the engines have learned your name.

The second play is getting your own pages cited, so your numbers, your framing and your vocabulary travel inside the answer even when nobody clicks. The experimental evidence on how is unusually direct. The Princeton-led GEO study, published at KDD 2024, tested nine optimization methods across generative engines and found that adding quotations, statistics and cited sources boosted a page's visibility by up to roughly 40 percent, with the best methods improving about 22 percent on position-adjusted word count and 37 percent on subjective impression. Keyword stuffing, the reflex of the click era, performed near the bottom.

Structure the work as one page per buyer question: a direct answer in the first paragraph, evidence with named sources through the body, and clean indexing so the retrieval pipelines can find it, since AI engines read search indexes and an unindexed page does not exist for them. Citations also collect the paradox bonus of zero-click environments: the cited brand gets the click precisely because it appears inside the answer, which is what the 35 percent AI Overviews click-through premium measures.

This is the half of the playbook Reachroller automates end to end. Tracking identifies the unbranded questions where engines answer without you, each lost question generates a publish-ready fix page with URL slug, title tag, meta description, schema markup and indexing steps, and a recheck after one to two weeks shows whether the stored answer flipped. One credit is one AI answer, a generated fix costs ten, and the tiers are on the pricing page, starting at $29 per month.

A playbook is only trustworthy if it survives its best objections, so here are the three strongest. First, the headline numbers have real measurement spread. The 48 percent AI Overviews trigger rate comes from specific tracked query panels, and other trackers report 13 to 25 percent depending on the queries they watch. If your category skews toward transactional or local queries, your buyers may see answer boxes far less often than the averages imply. The response is to measure your own questions rather than argue from industry averages, which is what a fixed question set gives you.

Second, the direct revenue attached to AI referrals is still small: around 0.18 percent of sessions in some panels. A skeptic can fairly say the channel does not yet pay for a headcount. The counterweight is that referral volume was never the main prize; the influence happens inside answers that emit no session at all, and the G2 finding that 69 percent of buyers changed vendors because of chatbot output is influence no session count will ever show. Budget for the shortlist effect, and treat the referral conversions as a bonus that keeps improving.

Third, none of this replaces SEO, and pretending otherwise is malpractice. Google's AI features cite from Google's organic index, ChatGPT Search grew out of Bing's, and an unindexed page is invisible to every retrieval pipeline. Zero-click marketing sits on top of indexability rather than beside it: the same crawlable, rankable page that classic SEO demands is the raw material AI answers quote. Keep the SEO plumbing, redirect the content strategy toward citable answers, and let the two disciplines share pages instead of budgets fighting each other.

Two habits keep zero-click measurement truthful. The first is self-reported attribution: a plain "how did you hear about us" field at signup or on sales calls. It is unfashionable and it works, because it is the only instrument that catches the buyer who was persuaded by three AI answers and arrived as direct traffic. When those free-text answers start saying ChatGPT, you have evidence no analytics platform will ever hand you.

The second is refusing false precision on the answer side. AI responses are probabilistic: SparkToro measured under a 1 percent chance that two identical ChatGPT runs return the same brand list, so answer-presence metrics are only honest as trend lines over repeated runs on a fixed question set. A single check is an anecdote, a screenshot is a sales prop, and a trend line is a measurement. Store the raw answers behind every number, because the first question a skeptical CFO asks is where the number came from.

Set expectations in quarters for the mention work and weeks for the citation work. Third-party footprint builds slowly; a published, indexed fix page can flip a retrieval-backed answer inside two weeks, and the recheck either shows the flip or it does not. Reporting both clocks separately keeps the program credible while the slow half compounds.

A minimal operating cadence fits inside a normal content team's week. Once: write 20 to 50 unbranded buyer questions and name three to six competitors. Weekly: run the question set across the engines you care about, store the answers, and log mention share and citation share per engine. Biweekly: pick the highest-intent lost question, publish a citable page that answers it, submit it for indexing, and schedule the recheck. Monthly: review the trend lines, retire flipped questions into a maintenance list, and promote new questions from sales calls and support tickets.

The manual version of that loop costs a few hours a week and proves the mechanics on your own brand. Reachroller compresses it to minutes: the runs, the storage, the branded-question exclusion, the per-engine scores, the generated fix pages and the rechecks all ship in the product, through official engine APIs only, with ChatGPT live today and Claude, Gemini, Perplexity and Grok built and rolling out. The honest caveat stands: it is a young product. The discipline it enforces is the one this article just described.

However you run it, start this quarter. Half the market is still absent from AI answers entirely, G2 puts zero-citation B2B tech brands at 51 percent, and answers are being formed and reinforced now for the questions your buyers will ask next year. In zero-click marketing the compounding asset is the engines' learned belief that your brand answers a problem. The earlier the lessons start, the harder they are for a rival to unteach.