The co-op model is the product.

A membership that was actually ownership.

Most loyalty programs are built on a straightforward premise. The customer transacts, accumulates points, crosses a threshold, and receives something back. What keeps them returning is the cost of leaving: the value they would forfeit by going elsewhere. It works, and it encodes a specific belief about the customer, that they are a revenue source to be retained through lock-in.

The co-op is built the other way around. A one-time membership fee that technically functions as a loyalty enrollment produces something structurally different, because it makes the customer a partial owner of the enterprise. Members do not just purchase more than non-members. They advocate at rates no transactional loyalty mechanic generates, because the structure has built genuine shared ownership rather than a discount relationship.

That created an unusual product problem. The belief was already correct and already encoded, in the ownership structure itself. The question was whether the commerce and marketing systems sitting on top of it behaved consistently with it. In practice they often did not. Personalization had grown up inside individual channels and divisions, each with its own data, its own targeting logic, and its own view of the person. Retail, digital, and the membership program could each describe a member accurately and describe them differently.

A member who had been with the co-op for years could still be addressed as an acquisition target. The structure said owner. The systems said segment. Every division had good reasons for the tooling it had built, and the fragmentation was nobody's decision in particular, which is the usual way a system drifts from what the organization actually believes.

The P&L and the platform, in the same hands.

• P&LA $250M+ revenue P&L against a $30M variable budget, inside the subsidiary's annual operational strategy.
• StrategyChannel strategy, forecasting models, and bottom-line operations as cross-divisional owner of scalable site and personalization systems.
• RoadmapThe technology roadmap for personalization and AI, owned alongside the revenue rather than beside it.
• TeamA 20+ person organization of managers, agencies, and contractors, led with empathy, empowerment, and servant leadership.

The combination is the part worth noting. Personalization programs usually fail at the seam between the team that owns the technology and the team that owns the number. Holding both meant a platform decision and a margin decision were the same conversation, and that the roadmap could be defended on the revenue it carried rather than argued for as infrastructure.

One view of the member, not six segments.

The work was to replace per-channel personalization with shared infrastructure: one view of the member that every division could draw on, so that recognition travelled with the person instead of resetting at each boundary. That is a less exciting sentence than a campaign, and it is the thing that made the campaigns coherent.

1. Personalization as infrastructure

   Enterprise personalization infrastructure that powered scalable 1:1 experiences across digital channels and divisions, rather than a targeting tool bolted onto each one.

2. Recognition over targeting

   The design question shifted from which offer converts this segment to what this member has already done with the co-op, and what would follow naturally from it.

3. Forecasting tied to the roadmap

   Forecasting models and channel strategy run against the same plan as the platform work, so capability landed where the revenue was actually planned.

4. One brand across divisions

   Shared systems meant divisions could speak to a member in their own context without fragmenting the brand into a set of unrelated relationships.

Two member level figures would strengthen this page and are not published here: [Member share of sales] and [Personalization lift]. Both are the kind of number that has to be cleared before it goes on a personal site.

Belonging is a structural decision, not a message.

Belonging is the human outcome product teams most consistently underestimate, because it is the one that cannot be added later. It is not produced by tone of voice or by a better welcome email. It is produced by structure: who the system says you are, and whether its mechanics behave as though it means it. The co-op had the structure. The job was to stop the systems from contradicting it.

The design decision that created belonging was not a marketing choice. It was a structural one: the co-op model is the product.

This is the same finding the Azure work produced from the other direction. There, a platform's capability outran the clarity people needed to reach it. Here, a structure already encoded the right belief and the systems around it had drifted. In both cases the gap was human, not technical, and in both cases it was invisible to the metrics the organization was already watching.

Which is the practical warning. Customers who belong to something behave differently from customers who are locked into something, and the difference shows up in retention curves, price sensitivity, response to competitive offers, and resilience when the company stumbles. Leaders who build these systems without examining the beliefs embedded in them will optimize the mechanics and wonder why the outcomes do not follow.