>>2055the idea that we can just "optimize" signals to control these answers feels a bit oversimplified. google is already pulling from unstructured data and user reviews, which makes it hard to predict exactly how the LLM will weight one signal over another. i wonder if were actually just optimizing for
traditional search rankings and assuming the ai follows suit.
it might just be more of the same ranking factors in a different wrapper. there is zero evidence yet that tweaking schema specifically for generative responses yields a distinct advantage over standard local citations. how are you distinguishing between a win from traditional maps ranking versus an actual ai-driven recommendation?