[Declanomous]

Thanks for backing me up. My experience Southwest is actually what made me think of that. I used my free drink coupon on a flight recently. The person sitting across from me ended up changing their order to a alcoholic beverage as well. It works just as well as purchasing a drink in any case, because you order before you present the coupon.

I suggested a card, such as a gift card, because I suspect watching someone swipe a card has a larger effect than watching someone hand over a printed coupon. It would also work for those back-of-seat infotainment systems that allow you to swipe a card to buy something.

I work in marketing for a non-profit, and I spend a lot of time figuring out how to get people to tell other people they gave to us. I'm sure this is marketing 101. The funny thing is that I probably overthink marketing experiments. I love pouring over data like Gardete did, but most of the other marketing people I know would just use A|B testing. Even though their experimental design makes me shudder, it gets the job done. From the perspective of a middle manager, it doesn't really matter if you understand the solution as long as it works. I think understanding why allows you to build on your successes and failures, but that's a self-serving bias.