Three Coalition Programs. Three Failures. One Pattern. | Resonance
NZ Flybuys shut down. AIR MILES Canada collapsed. Australia
Three Coalition Programs. Three Failures. One Pattern.
NZ Flybuys shut down. AIR MILES Canada got torn apart. Australia's Flybuys is structurally trapped. The coalition loyalty model keeps failing — not because the idea is wrong, but because the architecture creates the conflict.
Coalition loyalty — the idea that customers should earn rewards across multiple brands and redeem from a shared pool — is the most validated consumer demand in the industry. Nine in ten loyalty members want cross-brand redemption. The concept isn't theoretical. It's been attempted, at massive scale, multiple times.
And it keeps collapsing.
The usual post-mortem is operational: bad management, wrong market, poor timing. But when you line up the failures side by side, the pattern isn't operational. It's architectural. The centralized coalition model creates a structural conflict between the operator and the anchor brands — and the anchor brands always win.
Case 1: New Zealand Flybuys
The structural failure: When the coalition operator controls the program but doesn't control the customer relationship, anchor brands will eventually reclaim both. Woolworths NZ didn't need Flybuys — Flybuys needed Woolworths NZ. The dependency was asymmetric. The moment the anchor brand realized it was subsidizing a shared program without proportional data access, the math stopped working.
Case 2: AIR MILES Canada
The structural failure: Two layers. First, the centralized operator (Aimia) tried to extract value through breakage (expiration dates), destroying consumer trust. Second, the anchor brand (Air Canada) reclaimed its program because it didn't want a third party controlling its customer relationships and data. The coalition model created a landlord-tenant dynamic — and tenants with enough leverage don't stay tenants forever.
Case 3: Australia Flybuys
The structural trap: Australia's Flybuys didn't fail like NZ Flybuys or implode like AIR MILES. It's the "success" scenario — and it's still trapped. The program is geographically bound, physically dependent (POS-based), and invisible to the digital economy. Even at peak scale, it demonstrates the ceiling of the centralized coalition model: one country, physical retail, anchor-brand dependency. The Payback model in Germany shows the same pattern at even larger scale — 31 million members, 600+ partners, and REWE Group still exited to build its own program.
The Pattern
Line up all three cases and the structural failure is identical:
The common thread: centralized control creates asymmetric dependency. The coalition operator needs the anchor brands more than the anchor brands need the coalition. When the anchor decides it wants its own data, its own economics, or its own customer relationship — the coalition loses.
This isn't a failure of execution. Flybuys NZ ran for 27 years. AIR MILES was a household name for three decades. Australia Flybuys reaches half the country. These weren't poorly run programs. They were well-run programs on an architecture that inevitably produces the conflict that destroys them.
The consumer demand for cross-brand rewards is real. The failure mode isn't the idea — it's the architecture. Centralized operators who control the rails will always lose their biggest tenants.
What the Next Generation Requires
The coalitions failed because they were centralized, extractive, and physically dependent. The next generation needs to be the opposite.
Decentralized: Brands retain control of their customer data and economics. No landlord. No third party sitting between the brand and the customer. Joining the network adds capability without surrendering sovereignty.
Non-extractive: Revenue comes from engagement, not breakage. When the operator profits from breakage (as Aimia did with AIR MILES expiration), the incentive structure poisons the system. When the operator profits from flow — from credits being earned and redeemed across the network — incentives align.
Digital-native: APIs, not POS terminals. The next coalition needs to see a Discord message, a Shopify checkout, a Telegram reaction, a custom HTTP event — not just a barcode scan at a supermarket register. The digital economy is where DTC brands, communities, and creators live. A coalition that can't reach them is a coalition built for the last decade.
Additive: Every brand that joins makes every other brand's rewards more valuable. This is the network effect that coalitions promise but centralized architecture prevents. When value flows freely between participants — without conversion fees, without operator extraction, without geographic constraints — the flywheel actually spins.
The demand is validated. Nine in ten consumers want it. The failure mode is documented. The architecture that prevents it is clear.
The question isn't whether cross-brand rewards should exist. It's whether the infrastructure can be built without replicating the structural conflict that killed the last generation.