Reputation
Engagement metrics that actually matter for inbox placement
Major mailbox providers use machine-learning models that weight engagement signals: reads, replies, conversations, archives, deletes without read, and spam clicks.
Positive signals
Opens (with image fetch), replies, conversation continuation, message moved out of spam, message starred / labelled. These signals push future placement toward the inbox.
Negative signals
Spam clicks (worst), 'delete without read', unsubscribe (mildly negative — the sender just removed the channel), 'never seen this domain' (cold).
Why open rate alone misleads
iOS Mail Privacy Protection pre-fetches images, inflating open rate to ~30-50% on iOS-heavy lists. Use click-through, replies, and conversion metrics — opens are noisy.
Frequently asked questions
- Does sending less mail help?
- Often yes. Reducing frequency to engaged-only segments lifts every engagement metric and pulls placement back toward the inbox.