Managing the cost of fraud in incentive programs

No incentive program will exist without fraud.

The question is: How much fraud is acceptable in your incentive program? And how much is fraud prevention worth? Because it costs money to stop.

Keep asking these questions because there’s a very real tradeoff between the cost to minimize fraud and the expense of the fraudulent activity itself.

Here’s a real-life example:

Three sisters, all under the age of 9, were given a penny for every stick they picked up in their dad’s yard. I’m sure he thought it was a fool-proof incentive program.

Yet, later that day, he found them behind the shed breaking large sticks in half to increase their payday.

In case you haven't guessed, I was one of the sisters. And, despite the fraud, my father felt the extra $0.25 was worth it to have a stick-free yard.

For those operating incentive programs, fraud always exists. It’s up to each of us to decide how much fraud detection and mitigation is worth. And it could look different for each program, depending on your situation.

Understanding fraud in incentive programs 

Before evaluating the right balance between fraud and fraud prevention, understand what drives the need to eliminate fraud and the most common ways fraud occurs. 

The psychology of eliminating fraud 

When the idea of reducing fraud comes up, reflect on what’s driving the conversation. Typically, a single, memorable incident spurs the desire to try to eliminate fraud, and it’s fueled by hidden psychological factors:

• Loss aversion: Individuals in charge of incentive programs or retail organizations tend to feel losses more strongly than gains.
• Availability bias: Dramatic fraud cases are memorable and can influence perception.
• The "something must be done" fallacy: Emotional pressure to act quickly regardless of cost-benefit analysis affects decision-making.
  

The impact of psychological factors can be considerable, so knowing the factors helps you step back and evaluate if your emotions are getting the better of your judgement. 

That’s not to say you should never worry about fraud. We typically speak to clients at least once a year about these issues or as we find suspicious activity in parts of their incentive programs. The trick is to temper quick reactions with thoughtful consideration of how to proceed.

Thinking through fraud prevention tradeoffs

Remember, it’s not that you should strive to eliminate 100% of all your fraud. That’s unrealistic. 

There's a point of diminishing returns where additional fraud prevention measures cost more than the fraud they prevent. You need to find the right balance while recognizing psychological drivers.

Use this chart as a thought starter. The green diamond is where effort is worth the cost. Yellow signifies aggressive efforts that may or may not be worth it. Red is the mark of diminishing returns.

3 most common types of fraud

Fraud typically presents itself in three areas:

1. Data collection
2. Data validation
3. Redemption

We’ll go through each of them so you can recognize them in your programs, plus share tactics to detect and mitigate the risk of fraud in each area. 

Related: Learn the best practices for claims program management.

Detecting data collection fraud

Most fraud in incentive programs takes place in data collection within and around the following areas:

• Submission dates: Items sold outside the promotion-eligibility window
• Product information: Items listed as a different product than was actually sold
• Duplicate records: Sales records that have already been processed
• Quantities: Increases in the number of products sold on a real invoice
• Price ranges: Dollar amounts changed beyond a reasonable MSRP
• Location: Changing street or location naming conventions to resubmit sales or invoices
  

How to mitigate data collection fraud 

• Create data field validations: Check these during submission and include things like accepting certain types of data, limits on characters or amounts, immediate date evaluations for promotion eligibility, etc. We'll talk more about validation in the "Detecting data validation fraud" section.
• Limit the types of data allowed: Don't allow handwritten invoices or no quotes, for example, when participants upload a receipt, invoice or photo to validate submission.
• Collect additional meta data: Use data like who and where something was submitted (e.g., IP address, location, device and user) during validation to check for patterns or unusual submission circumstances.
• Create additional fields: Add fields for additional dates or product information to make it harder to edit validation documentation.
• Limit acceptable quantities: Create a maximum number of products or dollar amounts that participants submit at one time to keep the payout risk lower
• Deploy multi-factor authentication: Confirm identity as a simple best practice and one we always recommend.

Note: With every step of fraud mitigation there is a tradeoff to weigh. In data collection, the more fields you add, the more real-time field validation could increase your technology costs, the longer it takes and the more labor it requires for your incentive program participants to enter their sales information. This is likely to lower the amount of sales data you can collect and decrease partner satisfaction in your program.

Typically, we recommend less effort in this stage beyond adding one or two fields and some simple data-field checks in favor or more robust detection in the validation stage.

Detecting data validation fraud

Validation is the process of matching sales or data inputs against other information to make sure it’s accurate and reasonable. This step is where the best fraud mitigation can happen—catching the participant in the act, just like my father did when we were breaking sticks.

During the validation phase, we attempt to further whittle away at the types of fraud noted in data collection.

How to mitigate data validation fraud

• Normalize data: Remove spaces and expand or collapse abbreviations or acronyms. You can find these in address or naming conventions that limit payouts per household, dealership or person. Create data sets that are as consistent as possible.
• Duplicate checks: Run invoice or receipt numbers against prior submissions by all participants.
• Check unique identifiers: Confirm submission data against things like serial numbers, VINs, etc.
• Strengthen eligibility requirements: Add additional criteria for your participants like having a complete profile, having accepted T&Cs or tax information to earn a payout.
• Run comparisons and look for patterns: Check for numbers that go up in consecutive order that might indicate a series of fabricated sales data. Look for historical sales and payouts for things that are outside a normal deviation. Sometimes this shows up in the same people being paid out on every sale.
• Create a fraud risk list: Set a policy for how many strikes before being banned from the program.
• Conduct audits: Establish a threshold for dollars or numbers of submission that you’ll check manually to confirm any systematic errors in validation.

Note: Mitigating fraud at this step is the most useful, but it’s also the most time consuming (and, by correlation, expensive).

You’ll incur costs during setup, build and configuration to establish the checks described above. You’ll also likely accrue human operational hours to review flagged records. 

OCR technology and AI advancements help correctly identify patterns and data types, but the tools aren’t 100% accurate. This again is an area where you need to figure out how much it’s worth for you to find and eliminate fraud.

Detecting fraud during redemption

Fraud also occurs in redemption, when participants spend what they have earned.  The most common types of fraud at this phase include:

• Impersonation: Someone reaches out via program call centers or support methods to gain access to another person’s account and spend the earnings.
• Changed delivery locations: Swapping the delivery information from the real earners’ information to one that sends prizes or bank deposits to a different place.
• Time-based double redemptions: With gaps in data processing timing, sometimes a larger balance stays in place, enabling fraudulent double redemption. 

How to mitigate redemption fraud

• Have strong identity confirmation: Validate a participant’s identity in various ways. Profile information like addresses, other contact information and birthdates are helpful in confirming identity.
• Record prior shipment addresses: Make it easier to match prior shipment locations or addresses on the program profile against redemption data to flag transactions for further review.
• Monitor changing banking information: Collect meta data surrounding those changes to validate user identity or possible fraud.
• Check your payment processor’s balance transfer technology: Ensure your tracking mechanisms are fast enough to immediately remove funds based on redemptions.
  

Weighing the cost of fraud mitigation 

Remember, every incentive program will have fraud. It just depends how much risk and expense you're willing to tolerate while balancing the cost in human time and participant satisfaction.

Ask yourself these questions:

• How much is the fraud costing me?
• How much are we identifying today?
• What is a reasonable percent or amount of fraud for my program?
• Does fraud impact other product or inventory decisions I’m making and how much does that cost me?
• How much would additional fraud detection and mitigation cost me? Cost my participants in time and effort?
• Have I shared any existing high-risk fraud lists with my incentive provider?

Related: Discover how to set the right goals for sales incentive programs.

Incentive programs are worth it (even with the risk of fraud)

There will eventually be a point where you must accept the loss and risk because it’s not worth it financially to catch everything. But your incentive program is worthwhile even with the risk because good, behavior-based incentive programs improve performance metrics.

Increasing sales or desired behaviors, and collecting the data around sales, service and products has so much value. It’s the reason you built the programs in the first place.

Be sure to decide, alongside your incentive providers and program administrators, what your tolerance is. You’ve got this.

Want to work with an incentives expert to determine the right balance of fraud prevention for your program? Let’s talk.