Web Hosting Talk







View Full Version : Fraud Statistics


mpope
08-03-2001, 04:55 PM
Hello,

I've searched the archives, and haven't come up with a real definitive answer, so I thought I would open up a discussion!

My question is:
What percentage of your signups turn out to be fraudulent?

Also, What percentage of orders that are recieved with different billing addresses and IP addresses (when you do a whois at arin.net) are fraudulent?

I had a signup today that I would have approved without hesitation, but the IP address was from YugoSlavia, and the billing address was from Mexico. The person wanted to prepay for a year, but was also going for my cheapest plan. (Usually the fraudulent orders want to prepay for a year, but they also want the most expensive plan, so I thought this was odd...)

And finally - What percentage of orders from countries such as Russia, Yugoslavia, etc. are fraudulent? (I have yet to recieve one order from these countries that was not fraud.)

All comments will be greatly appreciated!
Thanks!

GordonH
08-04-2001, 12:07 PM
Hello
I have not had too many problems with people using Russian addresses but I have had problems with Russian residents using fake US addresses and stolen credit cards.

Basically we refuse all orders where the Card issue country does not match the address given in the order.
This is on top of AVS which might still pass the transaction.
We back this up with IP checks.
We also don't normally accept orders from Thailand, Indonesia, Malaysia, Palestinian territories because of the 100% fraud rate we had from these places in our first 6 months.

I have to say we have not had too many problems with Eastern Europe but I know many other companies have.

Interestingly I was just looking at our terminations due to Spamming and 90% of them are customers from the far northern states of the US and Canada (Canada especially).
We have never terminated a Californian for spamming even though they make up a sizeable chunk of our customer base.
I have no reasonable explanation for this statistical anomaly.

Gordon