Sift

Pros: Workflows functionality (is like the main actions an user can perform while inside our services platform) and Rules functionality (inside each Workflow one can set high number of options and criteria to execute actions (Blocking, Reviewing, etc.) these two features bring what we need in order to map fraudsters and behaviors alike, while at the same time they Machine Learning engine learns from us! Making everything we create inside the tools a lesson learn and then we can focus on the other kind of attacks, relying on the tool.

Cons: The reporting is so basic it contrast with the high functionality and advanced Machine Learning tool it needs to be improved. Here is where tools like Ravelin excel Sift Science by far!.

Overall: It can be improved, the support team take too much time attending customer requirements, and regarding the reporting, one instead focusing on the actionable data invest too much time wrangling with the data to get the information we need (asking via email for Sift asciende support, yes is frustrating). But despite these drawbacks, their Machine Learning approach is the best on the market as of today.

Pros: I like that it helps me find suspicious interactions between users. It does get me if a user has a network of accounts and if they are the same person by verifying if they are using the same device.

Cons: It is a bit tricky to use. There are so many options and so much to discover it can be a little bit overwhelming. But once you know what you're doing (kinda) you should be all set.

Overall: We get to know the activity of users on a website. Helps us know if there is some kind of iffy business going on. And I like that it is self learning. Meaning it learns from the decisions you make and helps you tag any bad users.

Pros: The scores we receive on transactions are usually quite accurate and fast, and they have helped us avoid many potentially fraudulent orders.

Cons: We have had some issues where we are unable to locate a transaction in the system to further investigate, even though we received a score for the order.