How to detect referral fraud with the Fraud center

The most common abuse is self-referral: one person playing both referrer and friend. The Fraud center flags referrals that look like abuse and lists the referrers behind them, so you only review the few that matter.

What you’ll set up

The signals that flag a referral, and the checks that block a discount outright.
A quick review of each flagged referrer: approve, reject, ban or exclude.

What happens when a referral is flagged

Bloop holds the reward for 14 days, tags the Shopify order BLOOP suspicious referral, and lists the referrer in the Fraud center. If nobody acts, the reward is approved when the hold ends, unless the referrer is banned or the order was cancelled or refunded. See order status.

When the purchaser’s email is exactly the referrer’s own (ignoring capitals, +tag and dots in Gmail), Bloop rejects the referral straight away instead of holding it.

Signal	Shown as	What it catches
Self-referral (email)	Self Referral	The same or a closely matching email, including the same name on another provider
Self-referral (matching last name)	Self Referral	Referrer and purchaser share a last name
Self-referral (matching address)	Self Referral	The order ships to the referrer’s default address
Same IP as referrer	Shared IP	5 or more referred purchases from one IP in 30 days, by default
High referral volume	High volume	7 or more referred purchases in 7 days, by default

The self-referral signals are the strong ones. A shared IP or high volume can also be a household, an office, or a referrer who’s simply popular.

How to set your detection rules

Choose the signals → Click Save
1. In Referral program, click the Fraud center tab.
2. Click Detection rules. It opens Settings → Referral → Fraud detection, which you can also open directly:
3. Under Fraud detection, tick the signals you want. For Same IP as referrer and High referral volume, set the number of purchases and the days to look back.
4. Under Fraud Prevention, tick IP address or Browser to block the discount outright (see below).
5. Click Save in the save bar:
All five signals start on. Switch one off only if it keeps flagging referrals you always approve. For one customer, exclude them instead.

Block abuse before it reaches the queue

The two Fraud Prevention checks don’t flag a purchase for review. They stop the friend getting the discount at all. Both start off.

Check	What it does
IP address	If the referrer and the friend share an IP address, the friend doesn’t get the discount
Browser	A referrer can’t follow their own link and claim the discount in the same browser

While a check is on, Bloop also flags an order placed from the referrer’s own IP or browser. Turn on Browser freely, it rarely blocks a real referral. IP address also blocks genuine referrals between people on the same home or office network.

How to review a flagged referrer

Check the evidence → Approve, reject, ban or exclude
1. In the Fraud center, click a referrer. Start with the lowest Days left to review: at 0 the reward is about to be approved.
2. Read why each referral was flagged.
3. Click View order to open it in Orders, then approve or reject it. Either closes the case here.
4. For a clear abuser, ban them. For a genuine customer who keeps getting flagged, exclude them from detection:

How to decide

Signal	What to do
Email or address match	Exact email, an alias on another domain, or the referrer’s address. Reject, and ban if it repeats
Last name only	Look closer. Family referrals are usually real
Shared IP only	With unrelated real emails, approve. It’s a household or an office
Shared IP + self-referral	Reject
High volume, varied	Varied IPs, emails and names: approve. This is a popular referrer
High volume, repeated	Repeated IPs or alias emails: reject and ban

Ban a referrer

Ban only a confirmed abuser. A ban disables their referral link and code and their unused reward codes, disqualifies their pending referred purchases, and stops new referrals counting.

A ban is not the same as Block contact in Contacts. Both stop the referrer’s link, code and pending rewards, but a block can be reversed with Unblock, and a ban can’t be undone in the app. Unblock doesn’t lift a ban made here.

Ban → Click Ban referrer from referring
1. On the referrer’s page, click Ban followed by their name.
2. Tick I acknowledge that the effects listed are irreversible.
3. Click Ban referrer from referring:
The app has no way to undo a ban, so a referrer banned by mistake stays banned. If that happens, contact Bloop support. When you’re not sure, use Block contact in Contacts instead, which you can reverse.

Exclude a referrer from detection

For a genuine customer who keeps getting flagged, for example a family ordering from one router, click Exclude suspicious referral detection on their page. Their future referrals stop being flagged; everyone else is unaffected. Enable detection turns it back on.

Fraud history

Fraud history logs every action: the date, the referrer’s email, the action, and whether it was the system or someone on your team. Search it by email, or open it from a referrer’s page to see only theirs.

The Fraud history log with date, referrer email, action and actor columns

Best practices

Review weekly. A flag nobody reads turns into an approved reward after 14 days.
Don’t reject on one weak signal. A shared IP alone is usually a household.
Exclude, don’t ban, a good customer. A ban can’t be undone from the app.

Referral orders - approve or reject flagged orders
Contacts - block or remove a contact
Preventing referral fraud without blocking real customers - the broader approach