Hacker News new | ask | show | jobs
by newobj 4925 days ago
The idea of a flat charge back fee is weird[1]. If I think about a store where I sell things for $1, it means if I exceed a 7% chargeback rate, my business is no longer profitable. Is that really all there is to it?

Edit: [1] Not weird as in weird on Stripe's part, but weird as from the perspective of someone who does not sell online, just seems like this one-size-fits-all approach probably makes a lot of low-margin selling really hard.

8 comments

Credit card companies charge at least $15 per chargeback, which makes dealing with small charges very unprofitable, especially with credit card thieves using small purchases as tests for valid credit card numbers.

A 7% chargeback rate is huge and would likely get you shutdown by your credit card processor.

7% chargeback rate?

Out of ~2500 orders I've had one chargeback and it was our fault for not processing the customer's refund sooner. I'm not open to disclose what we sell but I can promise you that I have some very angry customers.

Maybe you're confusing refunds with chargebacks?

I've experienced a lot of chargebacks, unfortunately. For some reason some gang of people from Vietnam keeps using my site to verify stolen credit cards work. Originally this all happened through a third-party service which I had no control over (I couldn't preemptively block suspicious transactions). I switched to Swipe and eventually saw the same fraudulent transactions coming through (one was 50 charges with the same card with a minute of each other).

Sick of it all I finally got minFraud setup and working with Braintree as the payment processor. So now, I use Braintree to authorize the card, then do a fraud check through minFraud. If the fraud level is low enough, I submit the payment for settlement, otherwise I void it.

Since putting minFraud in place I haven't had any chargebacks.

Why not do the minFraud check before auth? Won't that save you some auth fees?

minFraud is a very cost effective service. We have them as one of the solutions in our platform.

There's no fee for an authorization. That's just the check if the card is valid. The fee comes from submitting the charge for settlement.
The effort to handle a chargeback is not proportional to the amount of the charge, so I don't see why the fee would be.

If you read around on retail forums, you'll find that nobody would be able to get away with anything close to a 7% chargeback rate with traditional credit card processing, without getting their merchant account terminated with severe prejudice -- anything more than a percent or so raises alarm bells with the card companies.

7% chargeback rate is insanely high and I wouldn't be surprised if your merchant dropped you. Realistically anything over 0.5% and you're doing something wrong and that's even a magnitude higher than what I would shoot for.

Also, selling at $1 you're only going to make $0.67 or so per transaction (maybe less depending on how you get bilked in fees). You're going to have to do a lot of volume to make money selling at $1.

As others have pointed out, a 7% chargeback rate is pretty problematic -- the card brands (Visa and co) will intervene long before then.
It depends. Its not always just about the ratio of chargebacks but also volume.

So you might trigger some flags with the card associations if you say have the following

1) 1% chargeback ratio 2) 1% of revenue result in chargeback 3) Have like 50-100 (forget exact number) of chargebacks a month for like x consecutive months.

If you don't fix the problem, then bad things happen like additional fines....etc.

So for merchants with very low volume, a higher chargeback rate is sometimes permissible. Acquiring banks have their own risk assessment so they may allow it or may not.

Reading your edit -- yes, low-margin retail is -HARD-, for a lot of reasons that people don't really get until they try to do it.
Even "someone who does not sell online" (i.e. a low-margin retailer) has to pay a $15+ chargeback processing fee to the ISO/processor/networks.
If you sell 10 crappy $1 things a month it doesn't make sense. If you sell 10000 nice $1 things a month - it does.