Dunkin' Rewards: America Runs on Loyalty Programs With Correct Information

Dunkin' recently overhauled their Dunkin' Rewards loyalty program. Some regular customers are not pleased. Coffee rewards that used to be available after spending $40 on coffee now require at least $50 in spend. The program has also become more complex, with a different points earning formula, monthly boosters and a greater emphasis on food. With greater complexity comes greater risk of error.

I would describe myself as an occasional coffee achiever. I might stop for coffee early in the morning when walking the dog, but not that often. I currently have 257 points according to my app -- not enough for one free coffee.

Screenshot showing I have 257 points

So, imagine my surprise when I received this email. 

Points amount email
34,438 points = almost 69 free coffees!

The points value in the email is not only incorrect, it is outrageously wrong. Why would anyone hold on to 34,000 points?

In a prior role, I worked on a loyalty program mailing that included a mention of a customer's point balance. Here are some of the quality control steps I took:

1. Verified output data against source data
2. Verified lettershop proofs against both source data and output data
3. Requested manual review of customers with points balance more than two standard deviations above the mean against source data. 

Perhaps Dunkin' skipped step 3 above. What they could have done is ask themselves something along the lines of, "Who in the email communication appears to have enough points for more than 10 of the top points item? Should that really be the case?"

Lesson: 

When sending personalized information, verify the accuracy of all variable data including outliers.