Magento 2 Fraud Detection Extension

What does Fraud Detector do?

Fraud Detector checks every new order in your Magento store and gives it a fraud risk score from 0 to 100. A low score marks the order as Accept, a middle score sends it to Review, a high score marks it as Reject. Risky orders can go on hold automatically, so nobody ships them before a manager has looked.

The final call stays with your team. Each order shows why it got its score, check by check, and you decide whether to cancel it, release it or ask the customer for more details.

More than 400 Magento stores use it. On a typical store only about 1–2% of orders end up as Reject; most suspicious orders go to Review, where a person makes the decision.

Everything runs inside Magento. There is no external fraud service to sign up for and no fee per checked order.

Fraud Score at a Glance

A Fraud Score column in the Sales > Orders grid shows each order's score in green, orange or red, so the orders that need attention stand out without opening them. Orders are scored in the background (by Magento cron) a few minutes after they are placed, and any order you open is scored on the spot.

The Fraud Risk Score tab on the order page lists every check with its result, for example "Order placed from US, but shipping address is in Canada" or "Customer placed 2 refunded orders". From there you cancel the order, hold it or release it with the standard Magento buttons.

You can turn on Hold on Review status and Hold on Reject status. Then a risky order goes into Magento's real On Hold state, not just a label, and it cannot be invoiced or shipped until someone releases it.

Lock a Known Fraudster Out of Checkout

Flagging an order does not stop the same buyer from trying again tomorrow. Customer Lock does: a locked buyer is refused when they try to place an order, so no order, invoice or stock reservation is created.

• Every checkout path. The block works for logged-in customers, guests, and orders placed through the REST API or GraphQL (for example from a headless storefront or a mobile app).
• Lock from where you already work. Lock or release a buyer from the order's Fraud Risk Score tab, from the Fraud Check tab on the customer card, or for many orders at once with a mass action in the order grid. A Customer Lock column shows which orders came from a locked buyer.
• You choose what a lock matches. The customer account only, the account and email, or the account, email and IP address. IP matching is optional, because people on a shared connection could be blocked by mistake.
• Lock automatically. A custom rule can lock the buyer when an order matches it, alongside Accept, Review and Reject.
• A record of every lock. The Locked Customers page shows who placed each lock, when, why and from which order, plus a log of the order attempts that were refused. A separate admin permission controls who can lock buyers.

The buyer sees a neutral message that you write, at checkout and on their account dashboard. Customer Lock ships switched off, so nothing changes in your store until you enable it.

8 Built-in Checks and Their Weight

The score is a weighted result of eight checks. Each check can be switched off, and you set how much it counts towards the final score.

• Shipping and billing addresses: do the country, state, city, postcode and name match?
• Customer location: is the order placed from the same country as the billing and shipping addresses?
• Distance: how far are the shipping and billing addresses from the customer's IP location?
• Customer history: how many completed, refunded and canceled orders has this customer placed before?
• IP history: the same check for the IP address, over a period you set (365 days by default).
• Order amount: is the order total within the normal range for your store?
• Phone number: does it contain repeated digits or sequences typical of fake numbers?
• Customer email: does it contain the customer's name, or a keyword from your blacklist?

On the Fraud Risk Score settings page you also set the score ranges for Accept, Review and Reject, and preview the scores of recent orders before you save.

Custom Rules for Your Business

When a pattern matters to your store, write a rule for it. A rule checks order, customer, billing, shipping and payment data, including the shipping method, payment method, item count and whether the IP address is new, and sets the order to Accept, Review, Reject or locks the buyer. A Reject or Review rule wins over a lower score, and an Accept rule can clear an order the score sent to Review. For example, send every order shipped to a list of countries to Review.

Rules can also read the payment additional information that your payment module saves with the order. If your gateway stores the 3D Secure result there, a rule can accept orders where the liability has shifted to the card issuer, so they skip manual review. The user manual lists the available conditions with examples.

Orders that do not need checking can skip it: orders created in the admin panel, for example ones imported from a marketplace, and orders in statuses you select.

Fewer Chargebacks, Caught Before You Ship

A chargeback usually starts with an order that looked normal and was shipped. Fraud Detector catches risky orders before fulfillment: they get a high score, go on hold, and the buyer behind a confirmed fraud can be locked out of checkout.

It is not chargeback insurance: no one reimburses an order you decide to ship. It works alongside your payment gateway's own checks, such as address verification (AVS), the card security code (CVV) and 3D Secure, adding the signals that only your store has, such as the customer's order history and the distance to their IP location. Read more in our guide on how to prevent chargebacks in Magento.

Customer Location on the Map

When you review an order, a Google map on the Fraud Risk Score tab shows the shipping and billing addresses next to the place the order came from, based on the customer's IP address. An order placed in one country and shipped to another is easy to spot. The map uses your Google Maps API key; the IP location comes from a database built into the extension.

Social Profiles of the Buyer

For a deeper look at a high-risk order, the Fraud Risk Score tab shows links to Facebook, Twitter and LinkedIn profiles that match the buyer's name. Open them to see whether a real person stands behind the order before you make the final decision.