Blaze AI

As an AI marketing platform, Blaze relies on email for transactional, automated, and marketing communication alike — so when all three shared the same domains and IP infrastructure, a reputation hit on one type dragged down the others, including the transactional emails critical to the product experience. At the time, Blaze was sending over 2M emails/month through Iterable.

Blaze's case is a shared-infrastructure reputation collapse:

• No separation between email types: transactional, automated, and marketing emails all ran through the same domains and IP infrastructure, so reputation damage from one type spilled directly into the others.

• Critical mail exposed to marketing risk: transactional emails — the ones customers most need to receive — carried the same reputation baggage as bulk marketing sends, with no protective boundary.

• No dedicated warmup or migration path: without segmented infrastructure, there was no way to rebuild reputation for one email type without being dragged down again by the others.

The concrete result: spam placement at all major destinations, with inbox placement under 25% and a unique open rate of just 14–18% — a near-total loss of email as a reliable channel, transactional mail included.

MailSoar moved Blaze to a higher-quality Iterable IP pool and set up custom return-path, tracking, and image domains, separating transactional and marketing traffic by subdomain and IP.

The active database was cleaned before migration, marketing email was rewarmed starting from roughly 500 engaged users a day, and volume was scaled up gradually based on engagement and reputation — prioritizing recent openers and applying sunset rules along the way. Automated campaigns were migrated one at a time, sign-up consent and expectations were tightened, and DMARC was strengthened once every legitimate sending source was properly authenticated, all under continuous monitoring of reputation, inbox placement, bounces, and complaints.