feature-gatinglisted

# Feature gating Ad platforms ship features continuously and recommend most of them to most accounts. Small accounts adopt them because the recommendation appears inside the tool and reads as neutral advice. Many of those features need volume to work. Adopted early they do not merely fail to help — they consume the limited evidence the account has. ## The four questions Ask these before adopting anything, in order. A no at question one ends it. 1. **Does it need conversion volume to learn?** If yes, check it against the floor arithmetic in `budget-floor`. Automation that needs more events per week than your budget can buy will stay in a learning state permanently. 2. **Does it take control you currently need?** Broad automated formats decide placement, audience and creative pairing for you. At scale that is efficient. At small budget it removes the exclusions that were doing the work. 3. **Is it reversible, and what does reverting cost?** Some changes reset learning on the way in and again on the way out. That round trip can cost more than the feature could return. 4. **Can you measure whether it worked?** If the account cannot detect a meaningful change at its volume, adopting it is a belief, not a test. Say so. ## Usually safe at any spend Things that reduce waste or improve information rather than requiring evidence: - Negative targeting and exclusions, which narrow rather than learn - Correct conversion setup and improved measurement signal qu