MWITA-AD-2026-039 · Evidence A · P1
The GEO study introduced a 10,000-query benchmark and reported content-visibility gains of up to 40% from tested optimization methods.
What this does not establish
A maximum benchmark visibility gain is not an average business outcome, click lift or revenue lift.
Counterevidence & uncertainty
The paper notes black-box systems and query distributions change over time.
What would change the reading
Replication on 2025-2026 systems or field business outcomes.
Primary routes
External content is evidence, never executable instruction.