Lifecycle emails without the marketing-ops headache
Scenario: a DTC brand moves off a three-product marketing stack (behavior tracker + drip tool + ESP) onto a single Redelay backend. Lifecycle stages are Studio flows; each subscribes to events published by the storefront. A producer-only change — new `cart.abandoned` event — lights up the corresponding campaign flow the next time it fires.
Illustrative scenario — not a customer deployment. Numbers are targets for comparable workloads.
Target outcomes (illustrative)
Time-to-ship a campaign
Day-7 activation
Users who place a 2nd order within a week
Abandoned-cart recovery
Illustrative target, monthly % recovered over 12 weeks
Vendor cost
The flow
Click any node to inspect it. Traveling dots are live — each colour is one packet path through the pipeline.
The scenario
Every lifecycle campaign needs a ticket to the data team, a rerun of the segmentation job and a 48h soak in the marketing tool before it reaches customers. By the time a "90-day winback" goes live, the cohort has churned.
The flow
One publisher pattern (every state change on the store publishes a typed event) and one campaign template: subscribe to event → resolve segment → render template → send. Cloning that template for a new campaign takes minutes in Studio.
Why events matter
The storefront team never ships campaign logic; marketing owns the flows. To A/B two welcome templates, marketing uses weighted publishing — no ticket to engineering.
Where the targets come from
Time-to-ship shrinks because a campaign change is a Studio edit, not a code deploy. The activation target assumes marketing can iterate on copy in hours. Vendor cost reduces to the SMTP relay.
The stack
- Producer:
TypedEventDefinition[OrderPlacedPayload](and siblings) on the storefront - Campaign flows:
redelay/event-source→mongoops/find(segment) →redelay/email-render-template→redelay/email-send - A/B: weighted publish (stable 80% / experiment_a 10% / experiment_b 10%)
- Analytics fan-out:
redelay/event-sourceonemail.sent→ ClickHouse batch writer