"More than half of Google searches end without a click" is one of the most repeated statistics in marketing, and one of the most consequential if you believe it uncritically. The underlying research is real. The headline compresses away the details that determine whether it applies to your business.
What the studies actually count
The best-known zero-click analyses — from SparkToro working with Similarweb clickstream data, and separately from Semrush — measure clickstream panel sessions that begin with a Google search and record whether any onward click occurred. That methodology carries several properties worth naming:
The denominator includes every query type. Navigational searches for "facebook," unit conversions, weather checks, time zone lookups, and celebrity ages all count as searches. Many of these were never going to produce a click to a business site under any circumstances, in any era. They inflate the zero-click rate without representing lost commercial opportunity.
Refinement counts as zero-click. A searcher who queries, scans, refines, and then clicks on the second query registers a zero-click event for the first. From a user's perspective this is one successful research session. From the metric's perspective it is one zero-click search plus one clicked search.
In-Google clicks are often excluded from "clicks." Clicking through to Google Images, Maps, or another Google surface may not count as an onward click, even when the user reached a business through it.
None of this makes the research wrong. It makes the headline a poor guide to strategy, because the number that matters to a business is not "what share of all searches end without a click" but "what share of commercially relevant searches in my category end without a click."
The number you can actually measure
That category-specific number is available to you, approximately, without buying a clickstream dataset.
Take your Search Console query set, strip branded terms, and segment by intent. For each segment, compute CTR at your measured position. Then compare against your own historical CTR at the same position from two or three years prior. The delta is a reasonable proxy for click suppression in your category — it holds position roughly constant and lets the SERP environment be the variable.
Businesses that run this exercise tend to find a wide spread. Categories where the query is a question with a short factual answer show real, sometimes severe suppression. Categories where the searcher needs to reach a specific provider, compare priced options, or complete a transaction show much less. "How long does a root canal take" behaves very differently from "endodontist near me."
The strategic read
If click suppression is concentrated in short-answer informational queries, the response is not to abandon content. It is to shift the mix.
Down-weight pure definitional and short-factual targets. Content whose entire value proposition is answering a question that a SERP feature now answers in place has a structurally weak business case, however well it is written.
Up-weight queries where the answer is genuinely specific to a provider, requires comparison, carries cost or risk, or ends in a transaction. These resist interception because there is no single correct answer to lift into a box.
Reframe what the informational content is for. A well-structured explainer that gets summarized rather than clicked still does two things: it builds the topical foundation that helps the commercial pages rank, and it makes your brand the source that gets summarized. Which brings up the part of the debate that has changed most.
Being the cited source
When a SERP feature or an AI overview answers a query, it typically attributes. Attribution without a click still carries value — it puts your name in front of the searcher at the moment of the answer, and in AI search interfaces it is increasingly the only brand exposure available.
That value is real but hard to measure, and it should be described honestly rather than dressed up with invented numbers. "We cannot yet measure the revenue impact of being the cited source, but we can measure citation frequency, and we track it" is a defensible position. A confident conversion figure attached to an AI citation is not.
What not to conclude
The zero-click statistic is sometimes deployed as an argument that SEO no longer works. The data does not support that conclusion — it supports a narrower one, that a specific kind of SEO content has lost value while other kinds have not.
The practical response is a portfolio question, not an existential one. Measure suppression in your own category, move the content mix toward query types that still convert, and stop building pages whose only job is to answer something the SERP now answers for free.