01
The brief, in full
Customers heard from the business in pieces. Marketing, product and servicing each sent messages from their own tools, so a customer who had paused a loan application could receive a generic offer the same day — and an open complaint did not stop a promotion.
The brief: make every message aware of the last thing that happened, and make consent and regulatory suppression impossible to skip.
02
Our approach
We mapped the journeys from the customer’s side first: the moments that matter, the events that mark them, and the messages that should — and should not — follow. Then we joined app, web and servicing events into one profile, so the journey engine reads a single timeline.
Consent and suppression sit before the send, not after it. A message that fails a check never leaves the queue, and the reason is logged.
How it ran
W4W8W12W16W20W24
-
Weeks 1–4Journeys
Lifecycle journeys mapped with product, servicing and compliance in the room.
-
Weeks 3–10Events
An event schema for app, web and servicing, streamed into one customer profile.
-
Weeks 8–14Rules
Consent, preferences and regulatory suppression applied before every send.
-
Months 4–6Launch
Journeys switched on one at a time, each with a hold-out group to measure against.
03
What we built
An event-driven customer profile joining app, web and servicing events, with consent and regulatory suppression applied before any message is sent.
What it left behind
- 01Customer event schema
- 02Lifecycle journey maps
- 03Consent and preference centre
- 04Message audit log
04
What changed
Each journey is designed to start from the customer’s latest event, so a change in the preference centre reaches the next message, not the next campaign, and every message carries a record of why it was sent.
The measures below are read journey by journey, against a hold-out group. Results are shared only with the client’s approval.
How it is measured ›