The Economics of Card Loyalty: Where the Money Really Goes

Ask a card business leader how much the loyalty programme costs and you will usually get a clear number. Ask what it returns and the answer tends to get vaguer. That gap is worth closing, because loyalty is one of the largest discretionary costs in most card portfolios.

This article walks through the basic economics in plain terms, using simple, hypothetical numbers so the logic is easy to follow.

Where card revenue comes from

A typical card portfolio earns money from a handful of sources:

• Interchange, a small percentage of each purchase paid to the issuer.
• Interest on revolving credit card balances.
• Fees, such as annual fees, foreign exchange mark-ups and late fees.
• Indirect value, including deposits, cross-sell and the customer relationship.

Loyalty is supposed to grow these: more spend means more interchange, more engaged customers are more likely to revolve and hold other products, and a valued card is less likely to be cancelled.

Where loyalty money goes

Traditional rewards programmes typically carry four kinds of cost.

1. The reward itself

Every point, mile or cashback credit has a cost. In a points programme, the bank buys the reward, from an airline, a hotel or a catalogue supplier, when the customer redeems.

2. Programme operations

Platforms, catalogue management, partner integrations, fulfilment, customer service and fraud prevention all add up. These costs exist whether customers engage or not.

3. Marketing

A programme nobody knows about does nothing. Communication, campaigns and in-branch materials are ongoing costs.

4. Liability and accounting

Unredeemed points sit on the balance sheet as a liability. Breakage, the share that is never redeemed, reduces the eventual cost, but banks must estimate it carefully, and relying on it is an uncomfortable business model.

A simple worked example

Consider a hypothetical bank with 200,000 active credit cards and average annual spend of 5,000 per card in local currency. Total annual spend is one billion.

• If the programme returns an effective 1% in rewards, the reward cost is 10 million per year.
• Add operating and marketing costs, and the total can easily be meaningfully higher.
• If interchange averages, say, 1.5% in this market, gross interchange is 15 million.

In this example, a large share of interchange is flowing straight back out as rewards. That can be justified if the programme drives significant extra spend and retention. But many banks struggle to show that it does, especially for customers who would have spent anyway.

These figures are illustrative only. Real interchange rates, reward rates and costs vary widely by market, card type and regulation.

The incrementality question

The key question in loyalty economics is not "how much do we spend?" but "how much of the behaviour we reward would have happened anyway?"

A customer who already uses your card for everything and receives the same rewards as a customer you are trying to win is a cost with little incremental benefit. Generic programmes find it hard to tell the difference.

Targeted models change this. By focusing value on specific behaviours, such as activating a new card, trying a new category or returning after a period of inactivity, banks can direct spend where it actually changes outcomes.

How merchant-funded offers change the maths

Card-linked offers shift the funding model. The merchant provides the value, such as a discount, free item or upgrade, because the merchant wants the bank's customers through the door. The bank's role is to curate, target and distribute.

Going back to our hypothetical bank:

• The cost of the offers themselves is carried largely by merchants.
• The bank's cost is mainly the platform, targeting and marketing.
• Every redemption is measurable, so the bank can see which offers drive spend and activation, and which do not.

The result is a programme that can deliver strong customer value while keeping the bank's own reward cost low. Many banks then choose to reinvest some of the savings in a lean core rewards scheme, or in exclusive offers for premium segments.

What to measure

To understand loyalty economics properly, track a small set of metrics consistently:

• Active rate: the share of cards used in the last 30 or 90 days.
• Spend per active card, compared between engaged and non-engaged customers.
• Incremental spend, using control groups that do not receive offers.
• Reward cost per incremental unit of spend, not per total spend.
• Retention: card closures and dormancy among engaged versus non-engaged customers.
• Merchant health: how many merchants are active, redeeming and renewing.

The strategic view

Loyalty should be treated as an investment with a return, not a fixed cost of being in the card business. The right questions are: which customers, which behaviours, funded by whom, and measured how?

For many issuers, the answer points towards a mix: a modest, well-controlled core reward, combined with targeted, merchant-funded offers that do most of the day-to-day work.

If you would like to model the economics of a card-linked offers layer for your portfolio, cardoff.ai can help you build the case with your own numbers, from merchant funding to incremental spend.