See the fine print too. They have tiered service. I don't know why they are advertising their super plan on the front page as if its their whole offering...

"If you have Bolt's fraud indemnification, Bolt will fully cover the costs of and manage fraud-related chargebacks. But, you will still be responsible for Merchant-related chargebacks (damaged goods, goods not received, unhappy customers, etc.)"

"If you do not have Bolt's fraud indemnification, we will contact you via Email to inform you of the chargeback. Your merchant account will be charged that full order amount plus a $20 processing fee. Then, we will request the appropriate information to help you fight the chargeback with the card network. If you win the dispute, you will receive the full order amount credited back."

https://docs.bolt.com/docs/risk-fraud-and-order-review

We recently rolled all clients to the full offering. A couple legacy clients don't have the 100% fraud coverage, thus keeping it in the terms. But if you sign up with Bolt from this point forward, it bakes in the 100% fraud indemnification. Will probably even take that out of the terms soon.

So what percent of the transaction number is flagged as fraud, and how many is actual fraud? Losing a customer that is not fraudulent due to aggressive filtering is almost as bad as having a fraudulent one.

Good question. Addressed that here: https://news.ycombinator.com/item?id=16216455

Our top focus is order approval lift. Zero fraud is cool and all (and makes for a good posting title) but is really just an assurance that you should be comfortable as we start to approve way more orders.

How do you determine if a chargeback is “fraud related”?

Chargebacks have detailed reason codes which we programmatically sort. We also have a fraud analyst team that sorts through the non-fraud chargebacks to verify that they're not fraud related (some are mis-categorized by banks) before passing it off to the merchant.

So someone falsely claiming order not received is not covered. You don't cover friendly fraud.

Right now, we don't cover friendly fraud. The reason is that it's largely out of our control. We can't control how well merchants communicate with clients, how fast/reliably they ship, etc. So if we covered it, it lowers the cost of treating customers poorly. There will always be a bad actor here or there, but that's typically less than 1% of chargebacks (at least from our data).

Wow, you're saying 99%+ of chargebacks are claiming not authorized? Would not have expected that.

Not exactly. Friendly fraud is when a customer is a clear bad actor and, even though they receive the goods/services, they claim they didn't.

Merchant-fault is typically behind many of the non-friendly-fraud and non-stolen-card-fraud chargebacks. This includes late shipments, not adhering to return policies, damaged goods, etc.

That's effectively the same thing as zero fraud for anything that matters, if you never have to pay for it, no?

I mean, I guess if someone fraudulently signed up for a subscription, you'd lose the future expected monthly subscription fees or whatever.

But it's pretty much good enough. When I saw "zero fraud" advertised, my immediate assumption was that they paid for any fraud, and they were confident they had fraud low enough to do that and still be profitable. No other way to get truly "zero".

Spot on.