Newsletter Open-Rate Collapse After Frequency and IP Changes
Thread & Timber is a home goods retailer with 1.2 million opted-in contacts. Its weekly 'New Arrivals' newsletter averaged a 34% open rate and 4.1% click-to-open rate last year. In the last quarter, opens fell to 19% while click-to-open rose to 5.8%, and revenue per email dropped 12%. The team changed from one weekly send to three per week in January, added a 'Shop Now' preheader, and began using a new dedicated IP in February. Google Postmaster shows domain reputation at 'Medium'; spam complaints are 0.08%. Apple Mail represents 46% of opens. The design team wants to add a fourth send and use a single hero image; the CRM team wants to suppress anyone who hasn't opened in 90 days. The CFO asks whether the newsletter is dying.
1. Is the newsletter dying? Interpret the open, click-to-open, revenue, deliverability, and complaint signals. What would you check next?
Read the worked answer
Do not diagnose from opens alone. Opens fell from 34% to 19%, but click-to-open rose from 4.1% to 5.8%. That pattern is consistent with a smaller, more engaged opening population, not necessarily a dead list. Apple Mail is 46% of opens and Mail Privacy Protection makes opens unreliable: prefetching can inflate them, blocking can undercount them, and neither is a clean engagement signal. Revenue per email is the harder signal, down 12%. That can come from lower deliverability, higher frequency, weaker targeting, or a smaller engaged segment.
Check next. Segment delivered rate, bounce rate, unsubscribe rate, complaint rate, and revenue per 1,000 delivered, not per send. Compare the new dedicated IP period to the prior shared IP. Review Google Postmaster reputation, blacklist status, authentication, and seed inbox placement. Check whether the three weekly sends cannibalize the same segment and whether the fourth send would add new demand or just more pressure. A 0.08% complaint rate is near the watch threshold and should be monitored, not ignored. The medium domain reputation suggests a deliverability problem may be amplifying a content/frequency problem.
Conclusion. The newsletter may be fatigued, not dying. Before suppressing 90-day non-openers, test with click or revenue-based engagement, because suppressing on opens will remove Apple Mail readers who never trigger a reliable open.
2. Design a test to decide between adding a fourth weekly send and suppressing 90-day non-openers. Specify cells, sample, duration, primary metric, and guardrails.
Read the worked answer
Design. Use a randomized, stratified test across the 1.2 million contacts. Stratify by prior 90-day click engagement, revenue decile, and email client so Apple Mail does not dominate one cell. Four cells: (A) control, current three sends; (B) four sends; (C) three sends plus suppression of truly unengaged contacts; (D) four sends plus suppression. If a full factorial is too operationally complex, run two sequential tests, but do not change both variables at once without a factorial.
Sample and duration. Size for revenue per contact, not open rate. With 1.2 million, allocate at least 100,000 per cell after excluding recent purchasers and legal suppressions. Run for eight weeks to capture payback and frequency fatigue, with a pre-period baseline. Keep content and send times constant except for the variable under test.
Primary metric. Incremental revenue per contact over eight weeks, including downstream conversion. Secondary: click per delivered, conversion, unsubscribe, complaint. Guardrails. Stop or pause if complaint rate exceeds the agreed threshold, unsubscribe rises materially, or domain reputation drops. Suppression definition. Suppress only contacts with no click and no revenue for 180 days, not 90-day non-openers, because opens are unreliable. Analyze by engagement stratum; a fourth send may work for high-engagement contacts and fail for low-engagement ones. The decision should be segment-specific, not a blanket yes or no.
3. What deliverability actions would you take for the new dedicated IP and medium domain reputation before increasing volume?
Read the worked answer
Pause the volume increase. A new dedicated IP needs warming; adding a fourth send while reputation is medium compounds risk. Warm the IP with the most engaged segments first: recent clickers, recent purchasers, and confirmed opt-ins. Increase volume gradually and monitor placement by domain. Do not blast the full 1.2 million list to warm the IP.
Technical and list hygiene. Confirm SPF, DKIM, and DMARC alignment; check the sending domain and link domain; review blacklists and Postmaster Tools. Remove hard bounces and complainers promptly. Re-engage or sunset truly unengaged contacts rather than repeatedly mailing them. Use confirmed opt-in where appropriate and make unsubscribe easy.
Content and cadence. A single hero image is not automatically bad, but image-only emails can be clipped and give little text signal. Use a balanced text-to-image ratio, alt text, and a clear plain-text version. Keep the three-send cadence stable during warming; do not add the fourth send until placement and complaints are stable for at least four weeks. Run seed-list tests across Gmail, Outlook, and Apple Mail. Track delivered rate, inbox placement, complaint rate, and domain reputation weekly. If reputation improves, increase volume by engaged segment, not by list size. This protects the new IP and gives the frequency test a fair baseline.