Your CMO calls the loyalty program a crown jewel. Your CFO files it as a contract liability. Only one of them is reading the balance sheet correctly.
Brandmovers laid this out cleanly for finance leaders this year. Once loyalty points are established as a material right, the company must allocate a portion of each transaction’s price to the points, based on their estimated standalone selling price, which must be estimated from historical redemption data. In practice, a customer pays $100 and earns 100 points, and if those points carry an estimated standalone value of about a dollar, the company cannot recognize the full $100 as revenue at the time of sale. It must recognize roughly $99 as current revenue and defer the rest as a contract liability until the points are redeemed or expire.
One transaction, no big deal. But you do not run one transaction. At scale across millions of transactions, this deferred liability becomes a material balance sheet item. And here is the uncomfortable part: finance directors who are not involved in loyalty program design decisions are managing a financial reporting exposure they cannot see.
That last line is the whole problem. Marketing designs the earn mechanics. Finance eats the obligation. Nobody owns the full picture.
The liability is the least of your worries
Deferred revenue is boring and manageable. The uglier truth sits one layer down.
Half of all loyalty rewards go unredeemed. Fifty percent is the average, not the goal; healthy programs see 20 to 30 percent active redemption rates that drive measurable revenue. Read that as a relationship signal, not an accounting one. Half your members earned a promise from you and then walked away from it. That is not loyalty. That is inertia with a login.
Then the pool gets dangerous. The massive pool of unredeemed points, estimated at over $200 billion globally, creates an attractive target for fraudsters. Why? Because loyalty systems are the soft underbelly of the enterprise. Loyalty accounts often have weaker controls than core banking systems, making them attractive to fraudsters, with EY estimates suggesting roughly $1 billion in annual losses from loyalty fraud, and account takeover and redemption fraud common.
So your “asset” is a deferred liability on the books, a dormant relationship in the CRM, and a fraud surface in production. Add stale consent, the checkbox someone clicked in 2019 under terms you have since rewritten four times, and you have a privacy exposure sitting inside a marketing program. That is three risk categories wearing one loyalty logo.
Agents do not read your terms and conditions
Here is the part almost no one has priced in.
The next customer is not a person scrolling your app. It is an agent shopping on that person’s behalf. That agent does not open your PDF terms. It does not parse your quarterly email. It does not sit in your batch-built segment waiting for a campaign.
An agent needs a governed, machine-readable profile layer: entitlements, current point balance, redemption rules, consent scope, and eligibility, all exposed as structured data an autonomous system can query in real time and act on. Almost no retail loyalty stack has this. Your loyalty data lives in a human-readable console designed for a marketer, not an API designed for a machine.
If your program cannot answer an agent’s question in milliseconds, the agent routes around you. Your points become invisible. Invisible points do not drive redemption. They just sit there as liability, compounding.
That is where the Cost of Doing Nothing sharpens. CODN is not the accounting entry. It is the compounding gap between a program that carries risk today and a demand environment that will be mediated by agents tomorrow. Every quarter you defer the rebuild, breakage rises, consent ages, fraud exposure widens, and your agent-readability gap grows against competitors who are closing it. The liability is linear. The CODN is exponential.
Rebuild it or retire it. Doing nothing is not an option.
Run one honest test on your program. Three questions, no vanity metrics.
One: Can you prove incremental lift against a real holdout? In 28 loyalty audits run across 2024 and 2025, 19 programs had no control group or holdout cohort at all, meaning the reported ROI figure had no way to separate loyalty-driven lift from baseline demand. If you cannot isolate lift, you do not have proof. You have hope.
Two: Can you expose a governed, agent-readable profile layer with real-time balance, entitlements, and redemption logic? If not, you are already invisible to the fastest-growing buyer on the internet.
Three: Is your consent current, scoped, and provable per member? If not, you are one regulator away from a headline.
Three yeses means you have infrastructure worth investing in. Rebuild it as an intelligence layer, not a punch-card. One no means you are financing a liability and calling it a strategy. Then retire it with dignity, recognize the breakage, and redeploy the budget into something an agent can actually use.
The programs that survive the next 24 months will not be the ones with the most points outstanding. They will be the ones a machine can read, a regulator can trust, and a CFO can defend. Everything else is deferred revenue wearing a marketing badge.