Last updated September 2026: this piece separates two threats the industry keeps treating as one, with sourced stats for each and a decision rule for which to defend first.
Marketing teams keep asking the same question with two different meanings: “why is our organic traffic dropping.” Sometimes the answer is that Google still ran the search, but the answer box kept the click. Sometimes the answer is that the search never happened, because the buyer asked a chatbot instead.
Confusing the two leads you to fix the wrong thing. You rewrite meta descriptions for a user who never opened Google. Or you chase chatbot citations for a query where the SERP click was always the real prize.
Here is the split, with the numbers behind each side.
What is zero-click search?
Zero-click is a Google problem first. The SERP still exists, your page can still rank, and your content can still be the source Google draws from. The user just never has to leave the results page to get the answer.
This is not new. Featured snippets and knowledge panels have absorbed clicks for years. What changed is the scale: AI Overviews turned a minority behavior into the majority one, and the rate is still climbing, up from 60.45% in 2024 to 68% in early 2026 on the same SparkToro measurement.
What is zero-search?
Zero-search is not a Google problem at all. There is no rank to win and no schema markup that guarantees a mention. The chatbot answers from what it already knows, from a live web fetch it runs on its own, or from both, and your brand either surfaces in that answer or it does not.
This is the newer, faster-moving threat. A year earlier, only 29% of B2B software buyers started this way. The habit formed in a single year, and the same G2 research found buyers who trust a chatbot’s judgment return to it at every later stage of the purchase too, not just the first search.
The trend is not limited to the first touch. The same G2 report found that 71% of B2B software buyers now rely on AI chatbots for research at some point in their process, up from 60% in the prior wave of the survey. Zero-search stopped being a niche behavior. For expensive, considered purchases, it is on its way to becoming the default.
For a fuller picture of how “answer” replaced “rank” as the unit that matters, see the glossary.
Zero-click vs. zero-search, the stats side by side
| Zero-click search | Zero-search | |
|---|---|---|
| What happens | You search Google. An answer box, AI Overview, or knowledge panel answers the question before you click anything. | You never open Google. You ask a chatbot directly and get your answer there. |
| Scale | 68% of US Google searches ended without a click in early 2026, up from 60.45% in 2024 (SparkToro and Similarweb, 2026). | 51% of B2B software buyers now start research in a chatbot more often than in Google, up from 29% a year earlier (G2, “The Answer Economy,” April 2026). |
| Gets worse when | A Google AI Overview appears on the query: the zero-click rate climbs to roughly 83% on those searches (Semrush AI Overviews Study, 2025). | The purchase is high-stakes or repeat-frequency: buyers who already trust a chatbot return to it at every stage (G2, April 2026). |
| What you defend | Your presence inside Google’s own answer, not just your rank position. | Your presence inside the chatbot’s answer. There is no rank position to defend. |
| Where to watch it | The SERP, through AI Overview presence data. | The chatbot’s output, through prompt-level citation tracking. |
Why the industry conflates the two
Most articles use “zero-click” as a catch-all for any traffic loss tied to AI. That blurs two mechanisms with two different fixes.
Zero-click still starts with a search. Google runs the query, ranks the sources, and answers the question on its own results page. Your content can still be the cited source behind that answer, even if nobody clicks through to read it.
Zero-search never starts with a search. There is no ranking, no SERP, and no crawl-and-rank pipeline to win. The chatbot answers from its own retrieval, and your brand either shows up in that answer or it does not.
Treat them as one problem and you end up optimizing for the wrong page, or worse, for a SERP the user never visited in the first place.
Take a query like “best AEO tool for SaaS.” A few years ago, that query lived entirely on Google: type it in, open five tabs, compare. Today it resolves in one of two ways. Google answers it with an AI Overview: you still ran the search, and the box absorbs the click. That’s zero-click. Or a buyer opens ChatGPT and asks the same question directly, never touching Google at all. That’s zero-search. Both end in the same missed opportunity if you are not cited on either side, and neither is fixed the same way.
Which threat should you defend against first?
Match the defense to the query, not to the trend you read about most recently.
- Definitional and “how to” queries (“what is AEO,” “how to reset a router”): mostly a zero-click risk. High search volume, and the answer box handles it right on the SERP. Defend with direct-answer formatting, schema, and SERP-side monitoring.
- Comparison and evaluation queries (“best CRM for a 50-person team,” “X vs Y”): increasingly a zero-search risk, especially in B2B. Buyers now run these inside a chatbot before they ever open Google. Defend with citation-building and chatbot monitoring.
- Local and “near me” queries: mostly a zero-click risk today. Map packs and local AI answers absorb the click faster than chatbot habits have caught up for local intent.
- High-consideration purchases (software, professional services, anything researched over weeks): a zero-search risk. The G2 data is specific to B2B software, but the same pattern shows up anywhere a buyer already trusts a chatbot’s judgment.
- Branded and navigational queries (a company’s own name): low risk either way. The user already knows the destination.
The rule of thumb: if your traffic still comes from someone typing a question into Google, measure zero-click first. If your buyers increasingly skip Google and open a chatbot instead, zero-search is the threat, and it needs a different defense entirely, not better schema, but more citations inside the model’s own answer.
How to monitor each threat
Zero-click and zero-search live in different places, so you watch them with different tools.
For zero-click, you are still watching Google. Semrush tracks AI Overview presence alongside its standard rank data, so you can see which of your ranked keywords now trigger an answer box instead of a click.
For zero-search, you are watching a chatbot’s output instead of a search index. Ahrefs Brand Radar tracks how often, and how favorably, your brand gets mentioned inside AI chatbot answers, the prompt-side equivalent of a backlink report.
If you want the chatbot side in one dashboard instead of stitching together exports, Temso tracks citations and brand mentions across the major AI engines from $89/mo, with unlimited projects, users, and recommendations on every plan. See how platforms in this category get evaluated in our methodology.
Where to start
Zero-click and zero-search are different budgets, different tools, and often different owners on the same team. Measuring one does not tell you anything about the other.
Start by checking both numbers for your own brand this month: your AI Overview presence rate on the SERP, and your citation rate inside chatbot answers. Whichever gap is bigger tells you which threat to fund first.
Compare AEO monitoring tools that track both sides of this split at /rankings/aeo-tools.