We look at a lot of loyalty programs, and one number almost never shows up in the reporting. How many customers would have bought without the offer at all. The measurement finally got good enough to answer it, and the answer is uncomfortable. A large share of mass promotions was wasted on sales that were already coming.

The industry's answer is to fix the promotion. Smarter targeting, personalized offers, the same nudge for a smaller markdown. Spend less, get the same sale. It's a real improvement, aimed at the wrong thing. A smarter discount is still a discount. You're still paying for the behavior, just paying a little less.

The distinction that actually matters is simple. Bought behavior is a customer who only comes back for the offer. Earned preference is a customer who would choose you at full price. On a sales chart they look identical, the same order on the same day. On a P&L they're opposites. One costs you a discount every time. The other costs you nothing, because the customer already decided.

The most profitable customer, then, isn't the one who took the leanest discount. It's the one who didn't need a discount at all, the customer who reorders every few weeks without a coupon. That customer costs nothing to keep and returns the most margin, and no amount of targeting manufactures them.

Measure that. Not how efficient your promotions got, but how many customers keep choosing you without one. It's the only kind of loyalty that shows up as profit instead of cost.

A cheaper discount is still a cost. The profit is in the customer you never had to pay.

Want a second set of eyes on how much of your loyalty you're paying for, and how much you've actually earned?