Were they already loyal?

Members buying more often and spending more than non-members sounds like proof that the program works. But if your best customers were the first to join, you may be subsidizing purchases they would have made anyway – and using their existing LTV to justify the expense.

The question is not whether members are valuable; it is whether the program creates additional value after its costs.

Imagine a skincare brand reviewing its loyalty program.

The presentation leads with:

Members spend twice as much as non-members.

They also order more frequently, buy across more categories, and have higher LTV.

The recommendation seems obvious: expand the program.

Increase the rewards budget. Add another tier. Push harder on enrollment.

Then someone asks:

“What did these customers look like before they joined?”

The team goes back through the purchase history.

The customers who became members were already buying frequently. They already shopped across the range. They were already comfortable paying full price.

In this example, much of the spending gap existed before enrollment.

The program did not necessarily create those valuable customers.

Those valuable customers chose the program.

That does not prove the program failed. It means the comparison cannot answer the question leadership thought it answered.

Member value and program value are different numbers.

Member value describes the customers who joined. Program value describes the difference the program made. A customer generating $500 in annual contribution does not mean your loyalty program generated $500 – especially if that customer was already on the same trajectory.

Your most engaged customers have the strongest reason to enroll.

Someone who already expects to buy regularly has an obvious reason to collect points and unlock benefits. A one-time shopper has less reason to bother. Comparing those groups afterward can make the program look effective even without a meaningful change in behavior.

Redemption proves a reward was used – not that it created an order.

A customer redeeming a voucher might be making an additional purchase. Or they might be applying it to a basket they were already planning to buy. High redemption is not automatically high incremental value. You need to understand what changed beyond the discount being claimed.

Protecting existing loyalty can be valuable too.

The program does not have to make every customer buy more to justify itself. Keeping a customer engaged when they otherwise would have drifted away can be a meaningful outcome. But unchanged spending alone cannot prove that protection happened, just as higher spending alone cannot prove the program caused it.

The wrong question is:

“Do members spend more than non-members?”

The better question is:

“What would these customers have done without this program or benefit?”

That alternative outcome is what your dashboard cannot observe directly.

Looking before enrollment helps expose existing differences. Comparing customers with similar purchase histories makes the analysis more useful. Neither, by itself, removes the fact that people choose whether to join.

For a practical starting point, test a new benefit rather than trying to prove the entire program at once.

Keep existing entitlements intact. Randomly offer an additional benefit to one group of equally eligible members, while another continues with the normal experience.

Then compare the groups over the same period – including everyone assigned to each group, not just those who redeem.

You are measuring what the extra benefit changed, not how attractive the redeemers look afterward.

A loyalty program can recognize your best customers, strengthen the relationship, and support long-term LTV. But a higher-spending member base is not enough to show that the program made the business healthier.

Before increasing the rewards budget, separate the loyalty customers brought with them from the loyalty your program helped build.