Some time ago I got one such coupon, too, by attempting an unsubscribe, but it’s not automatic.
It seems fairly obvious these coupons are most definitely not the same for everybody, e.g. there were times when it seemed mostly everybody was given incentives, and others where only some customers had access to coupons.
Assuming coupons are not simply randomly given away, it would make a certain deal of sense to me, if Humble were tracking behavioral metrics and using them to decide who gets the coupon. For instance, they might get the idea that, say, Kiril is a good customer, so a coupon might get him back on the train, whereas their data might show that I’m not, and a coupon is not as effective in making me change my mind.
Random coupons also make sense when there isn’t enough solid data to make more informed decisions, and also for black box (what if) inquiries into customer behavior.
They could also be running experiments on the customer base, by taking a group with homogeneous metrics and partitioning it in sub groups, which are then offered different promotions (including no promotion at all), in order to track the performance and effectiveness of the promotion, its timing, etc.
Since it’s also fairly obvious that they are monitoring general performance metrics, the likelihood of coupons being handed out, is also realistically affected by how well a certain Monthly is doing compared to past data and expectations.