5 Best Practices for Preventing Refund Fraud in Food Delivery

Food delivery services have revolutionized the way we enjoy our favorite meals, offering both convenience and our favorite foods at our fingertips.

When everything goes according to plan, the customer orders food using their favorite food delivery app, a driver picks up the order and delivers it, and payments are exchanged successfully.

Of course, occasional hiccups are a natural part of a platform’s operations, and sometimes customers receive the wrong food or no food at all. In these cases, it’s usually the platform’s priority to ensure that the refund process is as smooth and simple as possible.

Problems crop up when less scrupulous users take advantage of fair refund policies for their own benefit.

Refund fraud and its impact on food delivery 

Refund fraud happens when someone illegitimately claims a refund for a good or service they successfully received. Also known as “friendly fraud,” or “first-party fraud,” refund fraud is a significant concern for food delivery apps and the restaurants they partner with.

Back in 2021, a popular Los Angeles restaurant was forced to close permanently after receiving so many fraudulent chargebacks and scam refund requests that they could no longer keep up with the demands of the business. 

More recently, in 2024, two men in France were arrested for a refund fraud scheme that cost their victim platform over €2 million euros. By taking advantage of the platform’s first order refund policy, the fraudsters were able to automate a system of account creation, taking orders on behalf of other users, and fraudulently requesting refunds for each order. 

One thing is clear: at scale, this type of activity has the power to do real damage.

Unfortunately, refund fraud can also be more difficult to identify and combat than other types of fraud. 

If someone uses a stolen credit card to order their food, that’s probably a more cut-and-dry case of fraud than determining whether someone is lying when they say their food order was missing an item. 

The last thing a platform wants to do is react negatively to legitimate refund claims, but this presents the challenge of separating abusive refund requests from legitimate ones. 

And, of course, even if the platform can identify an abusive refund request, there’s also the question of how they should respond. When it comes to stopping refund abuse at scale, prevention is the winning strategy. 

Five best practices for preventing refund fraud 

Combatting refund fraud presents the food delivery industry with unique challenges, but it’s far from impossible. 

1. Ask drivers to verify deliveries when possible 

Drivers can verify deliveries by taking a photo of the food at the customer’s door or by asking the customer to provide a code generated by the app to confirm both recipient identity and successful delivery. 

Though these practices can add a bit of friction to the delivery experience for drivers and customers, it’s one way to help reduce refund fraud numbers by limiting the number of misunderstandings surrounding missing or failed food deliveries. 

2. Identify and block repeat offenders 

Repeat offenses are the name of the game for any fraudster that wants to commit refund fraud at scale. 

Fraudsters know that if they request a refund on every single order, their account will be flagged, and they won’t get any further. 

That’s why fraudsters use multiple accounts (“multi-accounting”) to spread their refund abuse out more evenly. If a fraudster is dedicated enough, there’s no limit to how many accounts they’ll create to keep the scheme going.

In the €2 million euro example we cited earlier, investigators identified 137,000 accounts associated with the fraud. 

It’s critical that platforms have a way to identify individuals across devices and accounts. Shared signals like device intelligence and location behavior can help prevent the multi-accounting and ban evasion that enable serial refund abuse. 

If you can recognize that multiple accounts are all associated with one device, or that multiple accounts were all created from the exact same location, that’s a link that can expose a refund fraudster. 

3. Keep an eye out for collusion

Collusion between accounts on different “sides” of the platform—like courier, merchant, and consumer—can be a sneaky contributor to a refund abuse problem. 

Let’s take an example from Incognia’s experience working with a package delivery platform. 

The platform was seeing a problem with courier and consumer collusion. The consumer would order high value items like alcohol, and then claim the order never arrived. The courier would claim that the order was delivered normally, and then secretly keep the items for resell. 

In a twist, Incognia’s device intelligence revealed that both of these accounts belonged to the same person. 

Collusion doesn’t even require multiple people to be in cahoots—if one person has access to multiple different types of accounts, they can pull it off, too. 

It can be a blind spot for platforms, but collusion between accounts is one of the ways that fraudsters can really scale up a refund fraud scheme. This is another reason why shared signals to identify relationships between accounts is so important. 

4. Take advantage of risk assessment

Nobody wants false positives in their customer refund process. Platforms want to retain loyal customers, and in the refund process, the platform is coming from a position where a customer is already unhappy with the service they’ve received. 

This puts a lot of pressure on policymakers and customer service teams to lean in to more lenient refund policies, but not all refund requests have to be treated equally. 

Some users carry higher risk with them than others, and there are a lot of factors you can take into consideration when deciding how risky it is that a given user is a fraudster. 

For instance, have you seen their device before on other accounts? Their location? 

How old is the user’s email account? Is their phone number verified? 

Has the same payment method or credit card number been associated with fraud in the past? 

These are just a few examples of the attributes that can go into building a solid picture of who exactly it is asking your platform for a refund. 

5. Put your refund policy to work for fraud prevention 

Fraud response strategy varies from platform to platform, and not every platform will want to ban users outright, even if they’ve made a lot of refund requests. 

Sometimes you might have someone who commits refund abuse opportunistically, but who otherwise pays for their orders normally. 

Or maybe limiting false positives is a bigger priority than eradicating all potential refund fraudsters from the app. 

Whatever the reason, permanent user bans aren’t the only possible solution to catching a probable refund abuser. You can also use your refund policy to your advantage. 

When users reach the threshold of a typical number of refund requests, or they’re associated with other accounts that are guilty of refund abuse, they might become ineligible for further refunds until their account builds back trust with regular ordering behavior. 

In this case, platforms still have to be careful of multi-accounting—if fraudsters realize that policy is why their requests are being denied, they’ll make multiple accounts and cycle through them to stay below the radar. 

Refund abuse is a policy problem, but it’s also an identity problem. 

Food delivery platforms are committed to providing a good experience for their users, and sometimes that means providing refunds when things go wrong. However, when people abuse these policies, it can hurt the platform and the restaurants it partners with.

Using best practices like those outlined above, food delivery apps can safeguard the integrity of their refund process.