We had a loyalty programme. Customers had no reason to be loyal.

A large company with many product lines, a loyalty programme, and plenty of customer data — and members who had no particular reason to come back.

Each product division spoke for itself, so a customer who had bought one thing was offered whatever the business wanted to push, not what made sense for them. Collecting data about someone is not the same as knowing them: the messages were generic, and points on their own were not a reason to open an email, let alone buy again.

We built the programme around the household rather than the product catalogue — what a family had already bought, what stage they were at, what would plausibly come next. Every interaction was used to learn something, and that learning changed the next message and offer. Messages and audience splits were tested on a small group before the full rollout, and follow-ups went only to people whose behaviour suggested they wanted one. The result was fewer, more relevant messages instead of more of them.