What Is Search Generative Experience (SGE)?

How Search Generative Experience Worked

Search Generative Experience was Google’s first public attempt to put a generated answer at the top of its results page. When a query triggered it, SGE produced an AI “snapshot” — a short synthesized summary drawn from multiple web sources — displayed above the familiar list of blue links, with links out to the pages it drew on. It did not replace the results page; it prepended an answer to it, and let users ask follow-up questions in a conversational mode below the snapshot.

Crucially, SGE never shipped as a standard feature. Google ran it inside Search Labs, its opt-in program for testing early-stage Search experiences, so only users who joined the waitlist and enabled it saw the generated snapshots. That experimental framing was deliberate: it let Google gather feedback on accuracy, sourcing, and layout before exposing the feature to everyone. The same generative engine optimization fundamentals that matter for today’s answers — being retrievable by AI crawlers and supplying clean, citable passages — are what determined whether a page was drawn into an SGE snapshot.

Example of Search Generative Experience

The clearest way to see SGE is through its own timeline, which is fully documented in Google’s announcements. Google unveiled SGE at Google I/O on May 10, 2023, in a post titled “Supercharging Search with generative AI,” and rolled it out as an opt-in Search Labs experiment starting in U.S. English. For roughly a year it remained a test surface: users had to enable it, and Google iterated on how the snapshot looked, what triggered it, and how it cited sources.

The experiment ended by graduating, not by shutting down. On May 14, 2024, at the following year’s I/O, Google announced that AI Overviews would begin rolling out to everyone in the U.S., noting that people had “already used AI Overviews billions of times through our experiment in Search Labs” — the experiment being SGE. From that point the SGE name effectively disappeared from Google’s own language, replaced by AI Overviews for the launched feature and, later, AI Mode for the fuller conversational surface. So “SGE” today is best read as the historical label for the 2023–2024 test phase of what is now AI Overviews.

The transition also narrowed the feature. In the Search Labs era, SGE fired a generated snapshot for a very wide range of queries, including many where a summary added little; by the time it shipped as AI Overviews, Google had pulled the triggering back toward queries where a synthesized answer genuinely helped, and shrank how much vertical space the answer occupied by default. That is the practical reason old SGE screenshots mislead — the surface people optimized against in 2023 is not the one their content meets in Google Search today. What survived the rename unchanged is the underlying bargain: to be drawn into the answer, a page still has to be reachable by an AI crawler and offer a passage clean enough to lift and cite.

Search Generative Experience vs AI Overviews

They are not competing products but two stages of one thing: SGE was the Search Labs experiment, and AI Overviews is what it became once Google launched it to the public. Optimize for AI Overviews; treat SGE as the backstory.

Frequently Asked Questions

The Bottom Line

SGE was the proving ground, not the destination. It was the opt-in Search Labs experiment where Google tested putting a generated answer above its links, and after roughly a year it shipped that idea to everyone under a new name — AI Overviews. Understanding SGE is mostly useful as history: it explains where today’s AI answers in Google Search came from.