Most enablement teams measure AI rollout success with a confidence survey, reps rate how comfortable they feel with the new tool, and that number gets reported upward as proof of success. It's an easy metric to collect, and an easy one to be misled by. Feeling confident and actually changing daily behavior are two different things, and this panel gets into the harder, more useful question, alongside the parallel decision every team faces about whether to buy an AI tool or build one internally. Expect a candid, multi-perspective discussion rather than a single playbook. Key takeaways: - Why confidence scores are a weak proxy for real adoption - What to measure instead to prove behavior has actually changed - How the decision to buy or build affects adoption in practice