Databases Reference
In-Depth Information
buyer's name will appear on the merchant's Register app and all the
merchant has to do is to tap on the name.
Square wants to make it easy for sellers to sign up for their service and
to accept payments. Of course, it's also possible that somebody may
sign up and try to abuse the service. They are, therefore, very careful
at Square to avoid losing money on sellers with fraudulent intentions
or bad business models.
The Risk Challenge
In building a frictionless experience for buyers and sellers, Square also
has to watch out for the subset of users who abuse the service. Suspi‐
cious or unwanted activity, such as fraud, not only undermines cus‐
tomer trust, but is illegal and impacts the company's bottom line. So
creating a robust and highly efficient risk management system is core
to the payment company's growth.
But how does Square detect bad behavior efficiently? Ian explained
that they do this by investing in machine learning with a healthy dose
of visualization.
Detecting suspicious activity using machine learning
Let's start by asking: what's suspicious? If we see lots of micro trans‐
actions occurring, say, or if we see a sudden, high frequency of trans‐
actions, or an inconsistent frequency of transactions, that might raise
our eyebrows.
Here's an example. Say John has a food truck, and a few weeks after he
opens, he starts to pass $1,000 transactions through Square. (One
the spot in case it's a bad charge. Technically the fraudster—who in
this case is probably John—would be liable, but our experience is that
usually fraudsters are insolvent, so it ends up on Square to foot the bill.
On the other hand, if Square stops payment on what turns out to be a
real payment, it's bad customer service. After all, what if John is inno‐
cent and Square denies the charge? He will probably be pissed at Square
—and he may even try to publicly sully Square's reputation—but in
any case, the trust is lost with him after that.
This example crystallizes the important challenges Square faces: false
positives erode customer trust, false negatives cost Square money.
