Hacker News Reader

Comment by ada1981 | Hacker News Reader

ada1981

I’m a little confused on the scam.

1. Customer places order for $700 worth of Avacados.

2. Uber Courier accepts the order; goes to grocery store for which Uber has issued him a $700 debit card that can only be used at that grocery store.

3. Customer canceled the order, but Uber somehow doesn’t automagically cancel the debit card, so courier buys $700 of visa gift cards.

4. Somehow Uber doesn’t monitor that issued debit card balances on canceled orders are being spent and they are bleeding $$$.. until it goes over $1MM.

Is this right?

View on HN

The flow is

- Place order

- Purchase gift cards

- Cancel order

Uber sees a line item for $XXX, at say Vons, and that's all they see. Generally food items are non refundable/returnable, so they have to eat the loss on canceled orders. It's actually a pretty solid scam, the issue is hiding it at scale.

Several metrics probably make them stand out:

- ratio of canceled vs fulfilled orders; frequency thereof

- ratio of average cost of canceled order to that of other drivers

- ratio of canceled orders on the consumer accounts vs population

Had they stayed small time, they probably would not have been caught

Edit: a word

Yeah, but this is so easily prevented by making it so the order isn't cancellable after the order has been run on the debit card.

From my understanding the debit cards are dynamically turned on and off and authorized only for the amount of the order. I would guess that means that they have near real time ability to see if the order has been paid for by Uber. This is pure, lunatic, shitty product management. Can you imagine if Amazon allowed you to cancel your order after it shipped?

Easy in the technical sense, certainly. The issue moves to UX. Likely there's some metrics/calculus that suggest that eating the occasional customer caused loss is actually more beneficial to the bottom line than having the customer eat it.

I'd expect that having customers swallow the cost of late cancelations would likely alienate a large fraction of them; perhaps the calculus is such that the money saved doing this is actually less than the lifetime value of the average customer.

I've seen this said this a few times here in various ways, but the optimal amount of fraud is non-zero [0]. While it's feasible to eliminate most/all fraud in various arenas, it's not cost effective to do so.

[0] https://news.ycombinator.com/item?id=32701913

>Can you imagine if Amazon allowed you to cancel your order after it shipped?

Yeah that's a return and theirs entire forums dedicated to return scamming.

Back in 1999, ordered a book from amazon, it took over 5 weeks, asked for a resend (cause i thought the package was lost), got a got a resend... then both books arrived.

When the scammer, attempts to scam a scammer, who is then scammed by a another scammer; which scammer wins? Which scammer is at fault?

Curious on the downvotes here considering the detailed discussion that followed…

Reply on news.ycombinator.com