GLOSSARY
Synthesis
- Synthesis
- Synthesis is how an AI engine composes an answer from many sources and its own trained knowledge, rather than quoting one page. Recommendations are synthesized: the engine blends reviews, mentions and comparisons into 'the three best options'. You influence synthesis by being consistently present in the material it draws from.
When an engine answers 'who makes the most durable carry-on luggage', no single retrieved page contains that answer. The engine weighs review roundups, community threads, publication tests and whatever its training absorbed about the category, then composes a recommendation with reasons. That composition step is synthesis — and it's where shortlists come from, which makes it the mechanism with the most commercial consequence in AI search.
Synthesis rewards footprint, not any single page. An engine composing a recommendation is implicitly polling the sources it trusts: brands that recur across reviews, directories, communities and press get named; brands present in only one place get averaged out. This is why the off-site half of GEO — earned mentions, review depth, citable original assets, entity consistency — moves synthesized answers when on-page work alone can't. Engines read mentions, not just links, so even unlinked coverage counts toward the footprint.
Its counterpart is extraction — lifting a specific passage from a specific page — and the two respond to different work on different timelines. Extraction improvements can register in weeks; synthesis reflects accumulated reputation and moves over months, which is also what makes it defensible once won. Example: a niche protein-powder brand appears in two best-of lists, a dietitian's blog and steady Reddit threads; six months later 'best organic protein powder' answers name it unprompted across engines. Nobody optimized one page to do that. The footprint did it.