Start a new topic

Please can the Fraud Detection System be fixed / adjusted?

At present, the Fraud Detection system only works if the subscriber and any given referral of theirs have the same IP. 


The system "should" consider the subscriber and their referrals as "1 IP Set" so if there are any duplicate IPs in the Set, the subscribe is flagged by the Fraud Detection system.


I'm absolutely positive other members of UpViral have this same issue, as we've had it twice already, in just a week, on a very small scale of marketing, relatively speaking.


Please, take a moment to reflect on this idea, and hopefully you'll pass it up the chain straightaway for review.


Here's a real example of the problem, from a previous support request of mine #82427:


"Hello Support, 


I've contacted support before about Fraud Detection issues, and didn't really get anywhere, however now I have another 'crazy extreme' example that hopefully won't be ignored this time.


I have a subscriber:


jessa_alynie@yahoo.com 12 112.198.75.62 2017-11-12


...who has referred (12) subscribers to get the giveaway, and all (12) are on exactly the same IP address, although they're on a different IP address then she signed up with, 112.198.72.43


* Please could someone look into this, as this is absolutely, definitely, a subscriber that should be flagged by Fraud Detection, don't you agree?


** See Screenshot.


I look forward to your reply, thanks!


All my best,

Dale"


2 people like this idea

Hi Michelle,

Unfortunately, that's not always the case.

The same IP subnet (or as you call it an IP set) could be used by many people sharing the same building/geographical location or company. For example, all employees working at IBM could use different IPs from the same subnet. It is worth considering flagging them, but a manual human vetting would be required to assure the IPs are indeed fraudulent. 

Hi Kaloyan, 


Thanks so much for your reply and I really appreciate your comments :-)


I of course agree that it's not always going to be the case, even though it's always been so far from my observations, in the instances of the issues in UpViral to date (where the FDS hasn't flagged subscribers engaging in fraud to build their referrals). I am fairly new to the system though ;-)


It seems reasonable to avoid over-complicating this idea, as all I'm suggesting is "expanding" the specification of the Fraud Detection System to include any duplicate IPs relating to the subscriber and the total of their referrals. So if there are *any* duplicate IPs within these, the subscriber is flagged, for very quick-and-easy reference.


The manual human vetting process is vastly more efficient if such instances are initially flagged by the FDS, so you can clearly identify what actually needs human vetting, rather than looking at every subscriber's referrals, to see if they're engaging in fraud - and then, if so, manually blacklisting their emails and IPs (that's honestly not a good use of time).


Please consider making this adjustment, as I'm very sure it will be very advantageous to UpViral users, and most certainly address an ongoing hassle in the system that I'm running into repetitively.


Again, thank you Kaloyan!


All my best,

Dale


1 person likes this

Thank you for the details, Dale. I've submitted the request and we might be able to include that feature in Fraud Detection system. I'll keep you in the loop :) 


1 person likes this

Brilliant, thanks so much Kaloyan :-)


All my best,

Dale

Login or Signup to post a comment