Word-of-Mouth and Referral Marketing Blog

Announcing Fraud Center for ReferralCandy

referralcandy-fraud-center-alert-review-action-support-hero

Fraud is a big concern for many ecommerce merchants, which is why we’re proud to announce the launch of Fraud Center 1.0 for ReferralCandy. This was a highly-requested feature from our merchants, especially those with high transaction volumes or Shopify Plus brands.

Ecommerce fraud is becoming more prevalent and is expected to cost merchants US$20 billion in 2021. As merchants struggle with supply chain volumes and increased ecommerce orders, fraudsters have become more active, knowing that merchants have limited resources to investigate suspicious transactions.

Our goal with Fraud Center is to help our merchants combat fraud, by making it easy to detect, review, and take actions against suspicious customers and transactions. Our goal is to reduce headaches for merchants and make referral programs easier to operate.

What does the ReferralCandy Fraud Center do?

Fraud Center is a command center where merchants can monitor, inspect and take action against suspicious activity and fraudulent transactions.

Fraud Center enables four main actions to address referral fraud:

referralcandy-fraud-center-action-flowchart-suspicious-referral-fraud

  1. Alert: Identify customers with suspicious transactions
  2. Review: Examine and review the evidence of suspicious transactions
  3. Action: Disqualify transactions or ban advocates
  4. Support: Check a record of past actions taken for customer inquiries.

With these four actions, merchants have the power to address referral-related fraud. Fraud Center addresses two major types of referral fraud:

  • Self-referral: An existing customer pretends to be a new customer, using their own referral link to earn the friend offer and referral reward for themselves.
  • Exposed coupon code: An existing customer shares their referral link publicly on a coupon-sharing site. This allows complete strangers to make use of the referral incentive while allowing the customer to earn the referral rewards many times over.

Some merchants also restrict customers from posting their referral links in public, for example on Quora or Reddit. Merchants take exposed coupon codes seriously, as it represents a source of lost revenue.

Exposed coupon codes are a violation of a referral program

The goal of a referral program is to incentivize word-of-mouth, bring in new customers, and reward customers who become brand advocates. Most merchants would happily pay out rewards to a proactive customer, who shares their referral link widely and brings in new customers. Such a customer makes a great advocate for the brand.

But when a referral link or referral code is exposed on a coupon-sharing site, several things happen:

  • The merchant makes a loss: The referral code is used by customers who would have made a purchase otherwise. Since referral offers are generally better than welcome offers, the merchant makes a loss on the margins.Also, the referral rewards are being claimed by a customer who posted the referral code publicly, instead of a customer who tells their friends and family about the brand. The merchant is paying out rewards for fraudulent behavior.
  • No word-of-mouth sharing happens: The incentive is being claimed by customers who would have made a purchase anyway, instead of new customers who have been referred to the brand. The merchant is paying out the rewards without reaping the benefits of brand awareness or word-of-mouth sharing.
  • Merchant see skewed marketing data: A merchant may see an influx of new customers through their referral program, but this is inaccurate.

Exposed coupon codes are an abuse of the referral program, and have caused big headaches for many of our merchants. We’re pleased to share how Fraud Center can help our merchants address this problem.

Let’s take a look at how the four functions of Fraud Center enable merchants to combat fraud and retain revenue.

1. Alert: Identify customers with suspicious transactions

Fraud Center provides a place for merchants to identify all instances of referral fraud, as well as keep an eye on overall trends.

When you log into the ReferralCandy dashboard, the Fraud Center widget shows you the number of suspicious customers detected. These are customers participating in your referral program who have one or more referrals we regard as suspicious.

For security reasons, we won’t share the specific signals we use to identify potential fraud with referrals.

referralcandy-dashboard-widget--fraud-center-referralcandy

You can click [Review Customers] to head to the Fraud Center dashboard, or access the dashboard by going to [My Store] → [Fraud Center]:

menu-sidebar-access-fraud-center-referralcandy

Fraud Center dashboard

The Fraud Center dashboard provides a list of suspicious customers for your review, sorted by urgency. Customers nearing the end of their review period are placed at the top.

If no action is taken before the end of the review period, the pending rewards will be paid out to the advocate. Before that, you can take action to [Review] the transactions, and make a decision about whether or not to disqualify the referral and block the rewards being paid out.

fraud-center-dashboard-suspicious-customers--fraud-center-referralcandy

What the Fraud Center dashboard looks like for Shopify and BigCommerce merchants

For merchants on standalone platforms, the Fraud Center dashboard instead looks something like this. If exposed coupon codes are detected, a widget will also appear to alert you of this.

exposed-coupon-code-widget--fraud-center-referralcandy

You can click on the name of any customer to [Review] the profile of specific customers and their transactions.

Weekly report on suspicious referral activity

All merchants will also receive a weekly report on suspicious activity in your inbox. This helps you to keep an eye on the number of suspicious customers each week and be aware of trends.

suspicious-referral-activity-weekly-report-email--fraud-center-referralcandy

(For privacy reasons, the data emails are redacted in the report email but will be visible on the Fraud Center dashboard).

2. Review: Examine and assess suspicious transactions

Fraud Center identifies a customer as ‘suspicious’ if they have one or more referrals with fraud signals. These include:

  • Self-referral
  • An unusually high volume of referrals in a short time
  • Exposed coupon codes

While we are continually refining our system, there may be instances of customers incorrectly being flagged as suspicious. You will be able to review each customer’s fraud profile to make an informed decision with as much information as we can provide.

Reviewing fraud profiles of individual customers

Each customer’s fraud profile page highlights the suspicious referrals, the suspected cause of fraud, and a link to [Fraud History]. The profile helps you to quickly assess the customer’s pattern of activity and make a decision.

suspicious-advocate-fraud-profile--fraud-center-referralcandy

In this example, it looks like a false positive. Fraud Center flagged John Smith for an unusually high volume of referrals; but since there are only 2 referrals in 28 days, it is unlikely to be referral fraud (perhaps John Smith told both friends on the same day).

In this situation, we would likely select [It’s not fraud]. We cover the rest of the available actions in the [Action] section below.

However, you can also select [I’m not sure]. This opens up a window with options to email the customer for clarification or contact Customer Support. Our team will investigate and provide you advice.

not-sure-contact-support-referral-fraud-suspicious-transaction--fraud-center-referralcandy

If at any time you have questions about Fraud Center, please reach out to our Customer Success Team.

What does self-referral look like in Fraud Center?

The most common example of referral fraud is when customers attempt to refer themselves by creating another email address, or using an existing email address they have. Here is an example of an attempted self-referral, flagged by Fraud Center:

suspected-self-referral-fraud-profile--fraud-center-referralcandy

Fraud Center will highlight the suspicious string in the email address

In this example, Fraud Center has flagged this referral because the customer (whose name is Johnny) has referred a friend whose email address also begins with ‘Johnny’. This could be a coincidence, or Johnny may have used an alternative email address to claim the referral reward.

In this scenario, you can investigate further by selecting [View Invoice] or [View Shopify Invoice]. For example, if the shipping addresses for both orders match, it is likely a case of self-referral. You can then proceed to take [Action] to disqualify the referral or the customer from the referral program.

What if my customer is an influencer?

As a counter-example, you may have a customer who is an influencer or content creator who has cultivated a sizable audience. When they share their referral link, they might receive an influx of referral transactions, and be flagged by Fraud Center for a high volume of transactions.

Here’s an example of what their profile might look like:

content-creator-influencer-suspicious-customer-referral--fraud-center-referralcandy

Fraud Center has flagged this account for an unusually high volume of referrals. However, we can see the source of many of their clicks is youtube.com. In this instance, the customer has been creating content and sharing her referral link on YouTube.

You can investigate this further yourself to make a decision. You can select [It’s not fraud] to release her pending rewards.

If this is a known affiliate or content creator, you can also [Mark as trusted], so that Fraud Center excludes them from suspicious referral detection in the future. Some of our merchants also contact customers with the most referrals for partnerships.

What if an advocate has posted their coupon code online?

Fraud Center also searches coupon sites for exposed referral codes and identifies the owner of the coupon code. The fraud profile of such a customer looks like this:

exposed-coupon-code-fraud-profile--fraud-center-referralcandy

We highlight the coupon code exposed, as well as the coupon-sharing sites where we have detected their coupon code.

NOTE: We have been building our ability to detect and identify exposed coupon codes. However, coupon-sharing sites have an interest in obfuscating and hiding their codes, and so there is a possibility that your code may not be detected immediately. If you do see your coupon code in the wild, please share this with us at [rachell@anafore.com].

When purchases are made using those coupon codes, these transactions are flagged as suspicious as well. In this example, John Smith posted his coupon code on retailmenot.com, and it was used 14 times. Here’s what his fraud profile looks like:

exposed-coupon-code-dashboard-suspicious-advocate--fraud-center-referralcandy

You can visit the website to verify the presence of the coupon, and take direct [Action].

3. Action: Disqualify transactions or ban advocates

Fraud Center allows you to take action against customers with suspicious transactions. Let’s go back to the example of the customer suspected of self-referral:
suspicious-advocate-fraud-profile--fraud-center-referralcandy
After reviewing the customer’s fraud profile and activity, you can take several actions. Here is a flowchart of the actions you can take with Fraud Center:

referralcandy-fraud-center-suspicious-referral-fraud-mark-trusted-ban-advocate-actions

1. It’s not fraud (clear the transaction)

If you have reviewed the transactions but see nothing to suspect, you can select the [It’s not fraud] to release pending rewards to the customer.

This is also what will happen if the review period passes without action. This situation may arise when Fraud Center detects a false positive, for example, if a customer refers a friend who coincidentally has the same name.

Outcome:

  • Suspicious Transactions: Will be regarded as valid transactions.
  • Pending Rewards: At the end of the review period, pending rewards will be paid out to the customer.
  • Customer: Future transactions by this customer will still be examined by Fraud Center, and flagged if there are suspicious signals.

If you trust this customer, you can also choose to mark the customer as trusted.

2. Mark a customer as trusted

When a customer is Marked as Trusted, they will be exempt from suspicious referral detection by Fraud Center. All future referrals will automatically pass, and rewards will automatically be paid out at the end of the rewards period.
suspicious-referrals-mark-as-fraud--fraud-center-referralcandy
This is most useful if you know the customer in question, for example, if they are a power user, a friend of the brand, or an influencer you work with. You should only mark a customer as trusted when you are sure they will not engage in referral fraud in the future.

Outcome:

  • Suspicious Transactions: Will be regarded as valid transactions.
  • Pending Rewards: At the end of the review period, pending rewards will be paid out to the customer.
  • Customer: Future transactions by this customer will not be flagged by Fraud Center.

In most situations, you will want to mark suspicious transactions as not fraud. Only mark an advocate as trusted if you know them well and can trust they won’t engage in fraudulent activity in the future.

3. Disqualify referrals

If you identify a transaction as fraudulent, you can choose to disqualify the transaction from the referral program. From the customer’s fraud profile, you can click [Yes, it’s fraud]
ban-advocate-suspicious-referral--fraud-center-referralcandy.png
Clicking the “Yes, it’s fraud” button reveals the [Ban Advocate] button and also displays the suspicious referrals in question. Clicking [View & Delete Purchase] will open a page where you can examine the details of the specific referral transaction.

review-referral-transaction-suspicious-referral--fraud-center-referralcandy

From here, if you click [Disqualify this referral], the referral will not be eligible for referral rewards. Disqualified referrals will still be visible in the Fraud History, and a record will be kept for future review.

We recommend this option for merchants who want to disqualify a specific transaction, but still want to allow the customer to be a part of the referral program. For example, you may wish to send the customer a warning but still allow them the chance to earn rewards for future referrals.

[Delete this Purchase] removes the referral from our system’s records. This function is mostly used for testing, such as when merchants try out the referral program before launching it officially.

Outcome:

  • Suspicious Transactions: Transactions will be marked as ‘disqualified’ and regarded as an invalid referral.
  • Pending Rewards: Will not be paid out to the customer.
  • Customer: Existing purchases made by the referred friend will be disqualified, and future purchases made by the referred friend will not be eligible for referral rewards.

Note that this only applies to one specific referral; the customer will still be enrolled in your referral program and may be eligible for future referral rewards. If you wish to take more serious action, you can ban the customer.

4. Ban advocate

If the customer exhibits a pattern of repeated, fraudulent activity, you may decide to ban the customer from the referral program. You may wish to do so if:

  • The customer creates multiple accounts for self-referrals;
  • The customer has posted their referral links on multiple coupon-sharing websites
  • The customer has violated your referral terms of service
  • The customer has disregarded a previous warning

After clicking [Yes, It’s fraud], you can click the big red [Ban Advocate] button. This will permanently remove an advocate from the referral program.

ban-advocate-referral-fraud-suspicious-transaction--fraud-center-referralcandy

Outcome:

  • Suspicious Transactions: Transactions will be marked as ‘disqualified’ and regarded as an invalid referral.
  • Pending Rewards: Will not be paid out to the customer. Future rewards will not be paid to the customer.
  • Customer: The customer’s referral link and code will be deactivated; future customers attempting to use the code or link will not have a friend offer will be applied

4. Support: Check a record of past actions taken

All these actions are recorded within Fraud History, which gives a complete list of actions taken for your team to cross-reference.

Fraud History is useful in situations where a customer asks about pending referral rewards, or why they have been banned from the referral program. When the customer reaches out to Support, you can check Fraud History to see what action was taken and by which team member:

fraud-history-of-actions-taken--fraud-center-referralcandy

If actions are taken by the system automatically or by a member of ReferralCandy’s Customer Success Team, the action will be marked as by ‘ReferralCandy’.

Fraud History allows you to filter by the type of action. For example, to see all banned advocates, you can select [Action], then [Banned] to filter the list:
fraud-center-referralcandy-fraud-history-filter-by-actions
You can also search for actions related to a specific customer based on their email address. From John Smith’s fraud profile page, you can also select [Visit Fraud History] to review past actions taken against him:
fraud-center-referralcandy-fraud-history
This will also make it easier for your team to handle inquiries and coordinate across team members.

Fraud Center gives merchants options to address referral fraud

At ReferralCandy, our goal is to build features that make it easier for merchants to operate referral programs. Fraud Center is a much-requested feature from our merchants, and we’re proud to arm merchants with another tool to combat referral fraud.

If you’re looking for additional information, you can check out the following help articles:

For a full demo and inquiries, please book a meeting with our product designer Rachel, or reach out to our Customer Success team if you have any other questions or concerns.

Darren Foong

Darren works on Growth at ReferralCandy. He spends too much time reading fiction, performing improv, and doing things that don't scale. One day he hopes to be quite interesting.

Get our latest articles
in your inbox