Search Generative Experience, the name of Google’s experimental generative-AI search experience in Search Labs before AI Overviews launched more broadly.
SGE appears frequently in research, screenshots and SEO discussions from the experimental period. It included generated summaries and ways to explore follow-up information. Google described the experience as an experiment, so findings from a particular test should retain their date and context.
AI Overviews launched in 2024, but historical SGE observations should not be assumed to describe every later search feature. Interface layouts, availability and source presentation can change. Distinguish the experiment’s name from the product currently being measured.
For a business, the useful question is how a current search experience presents its information and sends visitors. An old screenshot can explain the history, but it is weak evidence for a present-day visibility promise or a universal click-through estimate.
Takeaway: Use current feature names and dated observations in reports; do not build a strategy around assumptions from a discontinued experiment label.
Example
A report says a brand appears in SGE based on a screenshot from 2023. The team preserves it as historical evidence and runs a fresh, clearly scoped AI Overviews review before claiming current visibility.
What to check
Record the exact query, market, date and feature observed. Keep experimental findings separate from current production behavior.
For more detail, see Google SGE learning features announcement.