Three Brands, New IPs, and a Peak-Season Sender Authentication Cutover
Meridian Travel Group runs three leisure brands from one Emarsys account: city breaks, beach resorts, and luxury cruises. Each brand sends from its own subdomain, but all share two new dedicated sending IPs. Volume is 5 million emails per month, and 70% is promotional. The parent domain publishes DMARC p=reject; the old ESP used a shared IP and a mismatched return-path. The group plans to switch all brands to Emarsys four weeks before peak booking season and wants to send the full 5 million in week one. The city-break list has 12% hard bounces and many inactive addresses; the cruise list is small but high-value. Brand teams want separate sender names and footers. The security team requires SPF, DKIM, and DMARC alignment for every brand, and the CFO wants no peak-season revenue loss.
1. Outline the sender-authentication setup for the three brands and the parent domain.
Read the worked answer
Authenticate each brand on its own subdomain so reputation and alignment are isolated, for example city., beach., and cruise. subdomains. For each subdomain, publish SPF that authorizes Emarsys and does not exceed DNS lookup limits, configure DKIM signing in Emarsys, and publish a DMARC record aligned with the From domain. Because the parent domain is p=reject, do not send brand mail from the parent; use the subdomains with relaxed or strict alignment as security approves. Use a custom return-path on the brand subdomain so SPF aligns with the visible From domain. Separate sender names and footers are content-layer choices and do not require separate IPs. Before cutover, send authentication test messages to external seed mailboxes and inspect the Authentication-Results headers for SPF, DKIM, and DMARC pass. The reasoning is that DMARC p=reject will block misaligned mail; authentication must be proven per brand, not assumed from the old ESP.
2. How should the two new IPs be warmed, given the four-week deadline?
Read the worked answer
Do not send 5 million in week one. Build a volume ramp by day, starting with the most engaged recipients-recent openers and clickers, and high-value cruise customers who are likely to engage-while suppressing hard bounces and long-inactive addresses. Send the same brand streams consistently from the same IP pool so receivers can build reputation, and throttle sends rather than bursting. Increase daily volume only if bounce and complaint rates stay within agreed thresholds and reputation signals are stable. Keep promotional and transactional streams logically separated if volume justifies it, but do not split two IPs into many small, cold streams. Monitor Google Postmaster Tools, complaint rates, and blocklist status daily. Because four weeks is tight, negotiate a reduced peak launch: warm the IPs with the best segments first and delay the broadest city-break promotion. The reasoning is that reputation is earned by consistent, low-complaint volume; a full-volume launch on cold IPs risks blocking during the highest-revenue period.
3. What should be done about the city-break list before any peak-season send?
Read the worked answer
Treat the 12% hard-bounce rate as a list-quality emergency. First suppress known hard bounces and invalid syntax, then run a re-engagement or sunset decision for addresses with no engagement over a defined window. Do not send a peak promotion to the entire inactive tail; segment it and send a low-volume reactivation test only after the IPs are warm. Implement bounce processing so repeated soft bounces are suppressed according to policy, and monitor complaint rates by acquisition source. Check that signup sources and consent records are defensible; if they are not, exclude those records. Seed-test the city-break creative and monitor the first batch before scaling. The reasoning is that a high hard-bounce list on new IPs signals poor hygiene to mailbox providers, and peak season is the worst time to teach them that signal. Cleaning the list may reduce short-term reach but protects the entire group's deliverability.