3 ms·
That's strange because when we do this, the person who initially paid usually ends up with more cash back after everyone else repays. Maybe this is a good way t
by robotcookies 11y ago
That's strange because when we do this, the person who initially paid usually ends up with more cash back after everyone else repays. Maybe this is a good way to measure the honesty of co-workers.
But I am the same way and hate trying to 'collect dues'. One approach would be to have everyone take turns paying for the entire meal and not worry about collecting individual payments. They do this in some countries regularly (I think Spain is one) - people take turns paying for everything so it works out in the end. Although I do recall a Spanish friend saying there was one person in her group who would habitually take longer to come up in the roster of "person paying entire meal".
It's an interesting idea for an app though. Instead of you nagging late payers, the app could send email reminders to them. There could be stickiness to an app that has some reputational knowledge of how individuals have paid up in the past. Sort of like a fico score but among friends or coworkers.
- ThrustVectoring 11y agoThat works only when you've got consistent participation in the group. It's a pain to track when people drop in or out. Just as fair is to randomly decide who pays for the entire thing - perhaps weighted based on what portion of the check they're responsible for.
- contingencies 11y agoOne approach would be to have everyone take turns paying for the entire meal and not worry about collecting individual payments. They do this in some countries regularly (I think Spain is one) - people take turns paying for everything so it works out in the end China famously operates like this. Interestingly, they call the western techniques of splitting the 'AA method', implying an equal split by participants.