AI Search

AI search spans a spectrum. At one end sit AI-native tools like Perplexity, built entirely around synthesized answers. At the other are hybrid experiences like Google's AI Overviews, where a generated summary sits above the familiar blue links. Most people now encounter both without thinking of them as different things.

Mechanically, what changed is the second step. Classic search retrieved documents and ranked them. AI search retrieves documents and then writes something new from them. The user reads one answer instead of comparing ten sources.

The commercial consequence is the zero-click result. When the answer is complete, the click never happens. Your content can be read by thousands of people while your analytics record nothing at all, because the reading took place inside someone else's interface.

That breaks the measurement model most teams still rely on. Sessions and average position were built for a world where visibility and traffic moved together. In AI search they routinely move in opposite directions.

AI search severs the link between ranking and traffic that every SEO metric was built on. You can hold your positions and lose visits, or lose positions and be cited constantly. Until you measure presence inside answers, you are reading dashboards that describe a version of search that no longer exists.

A publisher watches organic traffic to a well-ranked how-to article decline steadily while its position holds at number two. The cause is an AI Overview above the results that answers the question completely. The article is still being read — just inside Google's summary, by people who never load the page.