E-commerce Fraud Facts • Hacker News Board

Re: E-commerce Fraud Facts

#31

Post by pytrin » Fri, Sep 13, 2013, 1:30 AM UTC

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision.

I exchanged a few Email with their support previously, and there is no way to get real time fraud scoring. I would expect to receive in the response to a transaction event the risk score associated with it - something similar to what the Minfraud service does (which we use), but taking into account more factors since they collect more data via their Javascript API.

One can only hope they'll offer this capability in the future, and we'd be glad to try it out again.

Re: E-commerce Fraud Facts

#32

Post by rogerbinns » Fri, Sep 13, 2013, 3:40 AM UTC

ecopoesis wrote: post #27

joshuahedlund wrote: post #3

We've found that good indicators include: a large distance between billing and shipping addresses, a large distance between estimated IP location and billing address, large order size, using a free email like gmail/yahoo/hotmail (that's the smallest of the factors, but virtually all of our fraud orders use them). Even combining these and others with a threshold, it's still hard to reliably detect without too many fal…

At a previous gig we found the same basic factors. I wrote a quick script to iterate through all the available Weka[1] classifiers using our manually flagged data as a training set. Then I took the top 20 performing ones and used them on incoming orders in production. If more then half the classifiers agreed a transaction was fraud, we denied it. Though this seems a very blunt hammer (I'm not a machine learning exper…

What happens to the false positives - ie the ones you denied that were real? Do people get a way to prove they really are legit?

(Matt Cutt's blog claimed I was a spammer when I made a comment about two factor authentication - I have no idea why. It told me to email the administrator to get it accepted but of course provided no clues on what address to use etc so I didn't bother. I'm betting the anti-spam software is claiming a victory when it was actually a failure.)

Re: E-commerce Fraud Facts

#33

Post by wallywax » Fri, Sep 13, 2013, 3:49 AM UTC

pytrin wrote: post #31

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision. I exchanged a few Email with their support previously,…

E-commerce that deals with physical goods does have a time delay, so I think you're really trying to say that it's not a good fit for digital goods sales rather than that it's not a good fit for e-commerce...

Re: E-commerce Fraud Facts

#34

Post by wallywax » Fri, Sep 13, 2013, 3:50 AM UTC

jusben1369 wrote: post #14

larrys wrote: post #7

With domain name registration the factors that we have noticed that are almost certainly fraud orders are (in various combinations): 1) credit card payment is all lower case and/or obvious non understanding of how US addresses are formatted 2) domain name has "hack" or some foreign sounding word. Or is anything related to vietnam (get plenty from vietnam) 3) IP location doesn't match customers location 4) Multiple at…

Agree with Vietnam. As a % of fraud I'd say 70% is Vietnam. And few if any legitimate orders from Vietnam. In our experience.

And Nigeria and Romania.

Re: E-commerce Fraud Facts

#35

Post by ecopoesis » Fri, Sep 13, 2013, 3:59 AM UTC

rogerbinns wrote: post #32

ecopoesis wrote: post #27

Earlier quoted context omitted.

At a previous gig we found the same basic factors. I wrote a quick script to iterate through all the available Weka[1] classifiers using our manually flagged data as a training set. Then I took the top 20 performing ones and used them on incoming orders in production. If more then half the classifiers agreed a transaction was fraud, we denied it. Though this seems a very blunt hammer (I'm not a machine learning exper…

What happens to the false positives - ie the ones you denied that were real? Do people get a way to prove they really are legit? (Matt Cutt's blog claimed I was a spammer when I made a comment about two factor authentication - I have no idea why. It told me to email the administrator to get it accepted but of course provided no clues on what address to use etc so I didn't bother. I'm betting the anti-spam software is…

For any denial we popped a "technical issues, please call customer support" message. We found that fraudsters were far less likely to call then real customers.

Re: E-commerce Fraud Facts

#36

Post by aylons » Fri, Sep 13, 2013, 4:49 AM UTC

wallywax wrote: post #33

pytrin wrote: post #31

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision. I exchanged a few Email with their support previously,…

E-commerce that deals with physical goods does have a time delay, so I think you're really trying to say that it's not a good fit for digital goods sales rather than that it's not a good fit for e-commerce...

Thanks for pointing this. I was really confused with the parent post.

Re: E-commerce Fraud Facts

#37

Post by jasontan » Fri, Sep 13, 2013, 7:50 AM UTC

pytrin wrote: post #31

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision. I exchanged a few Email with their support previously,…

Hey Eran! We actually help quite a few e-commerce sites today. To be clear, 99% of our scores are updated and ready within a second of receiving the transaction. The fact that the response to a transaction does not include this score is a design flaw, and you're not the first to bring it up. We're working on a fully revamped API that should address this, and the revamped API will be launched in the next 2-3 months.

I do remember your emails from back in the day -- will followup offline.

Re: E-commerce Fraud Facts

#38

Post by pytrin » Fri, Sep 13, 2013, 7:54 AM UTC

wallywax wrote: post #33

pytrin wrote: post #31

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision. I exchanged a few Email with their support previously,…

E-commerce that deals with physical goods does have a time delay, so I think you're really trying to say that it's not a good fit for digital goods sales rather than that it's not a good fit for e-commerce...

Yes, you are right. I meant for digital goods eCommerce

Re: E-commerce Fraud Facts

#39

Post by pytrin » Fri, Sep 13, 2013, 7:57 AM UTC

jasontan wrote: post #37

pytrin wrote: post #31

We really wanted to like Sift, as we suffer from a substantial amount of fraud attempts (as most business who sell digital products). However, their model is not a good fit for eCommerce sites, and that's a shame - it seems to be built specifically for services marketplaces like AirBnb, where there is typically a time delay between payment and service provision. I exchanged a few Email with their support previously,…

Hey Eran! We actually help quite a few e-commerce sites today. To be clear, 99% of our scores are updated and ready within a second of receiving the transaction. The fact that the response to a transaction does not include this score is a design flaw, and you're not the first to bring it up. We're working on a fully revamped API that should address this, and the revamped API will be launched in the next 2-3 months. I…

Hi Jason,

Yes - that was the same claim I got from your support back then (score update within a second mostly) - but that breaks up a linear process into an asynchronous one (since you can't complete the purchase process immediately). Hope that makes sense - we will be looking forward to your API update.

Re: E-commerce Fraud Facts

#40

Post by _mulder_ » Fri, Sep 13, 2013, 12:46 PM UTC

jusben1369 wrote: post #14

larrys wrote: post #7

With domain name registration the factors that we have noticed that are almost certainly fraud orders are (in various combinations): 1) credit card payment is all lower case and/or obvious non understanding of how US addresses are formatted 2) domain name has "hack" or some foreign sounding word. Or is anything related to vietnam (get plenty from vietnam) 3) IP location doesn't match customers location 4) Multiple at…

Agree with Vietnam. As a % of fraud I'd say 70% is Vietnam. And few if any legitimate orders from Vietnam. In our experience.

Interesting, any reason for 70% exactly? Os is that just on your personal experience? I'd have thought Nigeria and East Europe would have been higher.