Coalition vs Proprietary: What Air Miles and Scene+ Teach Us
Air Miles didn't fail because coalition is a bad model. It failed because it lost control of the asset that mattered: the first-party relationship.
In 2022, Empire — parent of Sobeys and Safeway — left Air Miles for Scene+. In March 2023, LoyaltyOne filed for bankruptcy protection; BMO bought Air Miles for US$160 million, with roughly 10 million active members. In 2026, BMO announced the brand's end: Air Miles becomes Blue Rewards this summer, with miles converting to "Blue Points." The same week, Shell left Air Miles for Scene+. Meanwhile Scene+ — owned by Scotiabank, Empire, and Cineplex — passed 15 million members in Canada.
The easy reading is "coalition dies, proprietary wins." It's wrong. Scene+ is itself a coalition. What separates the two isn't the number of partners — it's who controls the customer relationship and the first-party data that flows from it.
Why Air Miles collapsed
The original Air Miles model was a pure issuance coalition: one shared currency, earned across many independent retailers, run by a third-party operator. The theoretical upside — a customer earns rewards everywhere — was also its flaw. No partner truly owned the member. Transaction data fragmented, the operator's brand wedged itself between the retailer and its customer, and the day a major partner walked, the network effect reversed overnight. Empire's exit triggered a spiral: fewer places to burn miles, fewer reasons to earn them, erosion of the base.
Why Scene+ holds
Scene+ is a coalition of a different generation. Anchored to a bank, it links card spend, grocery, pharmacy, entertainment, and now fuel into a single identity graph. Each partner adds first-party signal to the profile instead of diluting it. Shell joining isn't one more logo: it's a high-frequency category — the fill-up — that densifies what the program knows about every member. The modern coalition doesn't pool a currency; it pools data under a single, identifiable owner.
The real fault line
So the right axis of analysis isn't coalition versus proprietary. It's: who owns the identity graph and the relationship? A badly run proprietary program dilutes its signal just as thoroughly as a badly structured coalition. A coalition steered by a player that owns the data and the customer relationship can be more defensible than an underinvested in-house program. Air Miles lost the relationship. Scene+ centralized it.
What the Canadian operator should take from it
Before you join or build a coalition, ask the sovereignty question: at exit, do I get back a usable customer file, or just a points balance? If the program operator permanently wedges itself between you and your customer, you're renting traffic, not building an asset. PC Optimum — wired into DoorDash and anchored to Loblaw — illustrates the winning middle path: a single owner, centralized data, partners that enrich the profile without owning it. The lesson of the decade isn't "avoid coalitions." It's "never cede control of identity, whatever the model."