Generative search is search that uses large language models to compose an original, synthesized answer from retrieved sources, instead of ranking and displaying existing pages as links.
Why it matters
Generative search changes what "ranking" means. Classic SEO competes for position on a results page; generative search competes for inclusion in a single composed answer, usually via RAG, the engine retrieves candidate sources, then the model writes an answer citing a handful of them. For a SaaS founder, the practical consequence is that a well-structured paragraph on a trusted page (or an upvoted Reddit thread, Reddit is among the most cited domains in AI answers) can outperform an entire optimized site. Google's AI Overviews, ChatGPT search, and Perplexity are all generative search surfaces, each with its own retrieval sources.
How to use it
- Write passages that survive extraction: self-contained, answer-first paragraphs that make sense when quoted out of context.
- Cover the sources each engine retrieves from, your own site, review platforms, and the community threads that dominate retrieval for "best tool" queries.
- Measure inclusion, not position: track how often your brand appears in generated answers for your core queries (see share of model voice).
Related terms
Paul-Marie is the founder of Readyt, the Reddit growth platform for SaaS. He has generated 16K€+ in pre-sales in 2 months using nothing but Reddit, and now helps founders turn Reddit threads into their #1 acquisition channel.
