A meaningful share of developers now ask an assistant before they ask a search engine, and the answer names two or three tools. If you are not among them you are not in the evaluation at all, and you will not see it in your analytics, because the click never happened. Key learnings: - What appears to influence whether a model names your product, based on what practitioners have actually tested - The changes to docs, comparison content and third party presence that seem to move the needle - How to monitor your presence in AI answers without standing up a research function The session covers what is known, what is still guesswork, and what is worth your time this quarter. Expect honesty about the limits of the evidence rather than a tidy five step framework.