“They caught that our Friday flash pushes were cannibalising Saturday transactional opens. The first fix was a calendar change, not new copy.”
Helen Okonkwo — CRM lead, Midlands retail application
Client stories
What changed after the review
Teams describe specific cadence, copy, and quiet-hour decisions — not star ratings. Names and sectors are shared with permission.
“The cadence clinic was useful, though I wished we had blocked a full day. Ninety minutes went quickly once we opened the send map.”
James Whitfield — Product manager, membership app
“Segment briefing showed our ‘highly engaged’ cohort had the worst opt-out rate on promotional pushes. We stopped treating them as an unlimited audience.”
Priya Nair — Retention analyst, travel booking service
Extended story: quiet hours for a Birmingham retail app
A West Midlands retailer asked us to review eight weeks of push notification engagement after support tickets mentioned late-evening alerts. Aggregate open rates looked steady; tap-through on promotional stories had slipped for three consecutive Fridays.
During the engagement review we separated transactional reminders from promotional stories and plotted both against quiet-hour settings. Promotional sends at 20:30 overlapped with the largest share of “delivered, not opened” events. Moving those stories to 16:00 for a two-week trial restored Saturday morning visits to saved items without increasing total weekly volume.
The experiment backlog left two further tests: shortening titles on stock alerts, and pausing promotional pushes for seven days after a new consent. The retailer ran the first test with their existing messaging vendor; we joined one follow-up call to read the results.
Extended story: consent week for a fitness booking service
A UK fitness booking application saw strong welcome-message opens and a steep drop in permission retention by day ten. Our segment briefing isolated first-week cohorts and counted stacked tips, class reminders, and partner offers on the same devices.
The team capped promotional pushes for seven days after consent and kept class reminders only. Week-two permission retention improved enough that they postponed a planned increase in send frequency. Copy experiments moved to the backlog rather than launching during the fragile consent window.