Skip to content
MouseCat
Trends and Rules

Stop new fraud trends before the first chargeback

MouseCat spots new fraud and financial crime risks the day they start, confirms them without waiting for labels like chargebacks or analyst dispositions, and hands your team a backtested rule that shuts them down. The losses never get the chance to pile up.

01 / Discover

Trend detection

Spot new fraud trends the day they start

As soon as your transactions arrive, MouseCat compares them with your historical traffic, looking for groups of similar activity that are suddenly growing. It checks combinations of payment, account, device, and behavioral details, so a new attack stands out as its own pattern on the day it starts instead of hiding in your overall numbers.

Illustration of one new fraud trend over three weeks. With MouseCat, the trend is caught and a rule is live on day 1, and losses stop there, at about $1,100. Without MouseCat, losses climb until the first chargeback arrives on day 9 and a rule finally goes live on day 14, after about $192,900 is lost, and chargebacks keep piling up after that. The numbers are illustrative.

Know which new trends are fraud before any chargeback arrives

A sudden spike isn't always fraud. A promotion that's working looks like a spike too. MouseCat's investigator works through the cases inside each new trend the way your analysts would, checking the accounts, payment details, devices, and behavior, and tells you which trends are fraud or money laundering and which are healthy growth. It doesn't need a single chargeback or dispute to do it.

Illustration: two new trends appear in the same day of orders. MouseCat investigates both. A promo code spike is made of long-standing customers and lines up with a marketing email, so it is not fraud. A spike of orders from brand-new accounts uses cards in other people's names and a handful of shared devices, so it is a fraud ring and a rule is drafted. Neither trend has any chargebacks yet.

02 / Mitigate

Rule generation

Turn each fraud trend into a backtested rule

MouseCat writes a rule that targets the fraud pattern and backtests it against your history, so you see how much of the attack it stops and how few good customers it touches before anything changes. You can also start a rule from a few example cases or a plain-English description.

Illustration: a draft rule built from the fraud trend, declining orders from accounts under a day old when the cardholder name doesn't match the account and the device is shared by three or more accounts. The backtest shows it stops the attack, declining all 11 of today's fraud-ring orders, and leaves good customers alone, declining just 2 of 412,000 good orders in the last 90 days. It is written for the customer's rules engine and ready to start in shadow mode.

Shut the fraud trend down the same day, with your team's sign-off

MouseCat connects to your existing rules engine and puts the new rule into shadow mode, where it runs on live traffic without blocking anything. Your team sees exactly what it would have caught, looks at it with a skeptical eye, and turns it on with one approval. The fraud trend is stopped the same day it started.

Illustration: on the day the trend started, MouseCat spots it at 8:41 AM and confirms it is a fraud ring at 8:52. A rule is proposed at 9:12 AM and starts running in shadow mode at 9:15 without blocking anything. By 11:52 AM it would have declined 4 orders, all in the fraud trend. The fraud team checks those results and approves the rule at 12:09 PM, and it goes live in the rules engine at 12:13 PM.

Stop the next fraud trend on day one

See how MouseCat spots a new fraud trend, confirms it without waiting for chargebacks, and hands your team a rule that shuts it down.

Book a demo