Riskified Analysis Reveals 1 in 4 Refund Dollars Is Abusive; Introduces "Dynamic Returns," a New Policy Protect Feature, to Safeguard Revenue While Increasing Customer Satisfaction

Key Terms

AI-powered technical

"AI-powered" describes technology that uses artificial intelligence to perform tasks, make decisions, or analyze information automatically. It’s similar to having a highly skilled assistant that can learn from data, recognize patterns, and improve over time, helping to make processes faster and more accurate. For investors, this means better insights and more efficient operations, potentially leading to smarter investment choices.

item not received (INR) technical

A buyer claim labeled “item not received (INR)” is when a customer tells a seller or payment provider they paid for goods or services but never got them, similar to paying for a meal that never arrives. For investors, a high or rising INR rate can signal shipping problems, fraud, customer-service failures or increased chargeback costs, all of which can reduce revenue, raise operating expenses and harm confidence in a company’s sales and fulfillment processes.

customer satisfaction score (CSAT) technical

Customer satisfaction score (CSAT) measures how happy customers are with a specific product, service, or interaction, usually captured by asking customers to rate their satisfaction on a simple scale. For investors, CSAT acts like a thermometer for customer goodwill: higher scores suggest stronger repeat purchases, loyalty and positive word-of-mouth that support revenue and valuation, while falling scores can signal rising churn, costlier customer acquisition and future sales pressure.

Dark Web technical

The dark web is a hidden part of the internet that standard search engines and browsers don’t show, accessible only with special software and settings that mask users’ identities. For investors, it matters because stolen customer data, leaked corporate documents or illegal marketplaces found there can lead to regulatory fines, cleanup costs, loss of customer trust and sudden drops in a company’s stock value—like a hidden back alley that can damage a storefront’s reputation and finances.

network-level signals technical

Network-level signals are patterns and measurements gathered from the flow of data across a company’s or service provider’s technical infrastructure—such as traffic volume, device connections, and routing events. Like watching cars on a highway to judge how busy a city is, these signals give investors a real-time view of user activity, service reliability and potential fraud or outages, helping assess growth, operational health and risk without relying solely on reported financials.

NEW YORK--(BUSINESS WIRE)-- Riskified (NYSE: RSKD) a leader in ecommerce fraud and risk intelligence, today released a research analysis highlighting a growing customer experience dilemma: As merchants tighten controls to fight a surge in return and refund abuse, they are inadvertently creating a more restrictive and frustrating experience for their best customers. To help retailers mitigate this challenge, Riskified has introduced a new feature in its Policy Protect solution, Dynamic Returns, AI-powered return decisions that adapt in real-time based on customer eligibility.

Riskified's 2024 analysis of over a million refund claims found that 1-2% of total order value measured in sales dollars was requested back as refunds, with nearly 1 in 4 dollars claimed being abusive. In response to rising abuse, many retailers are implementing restrictive tactics like flat return fees, shorter return windows, and delaying refunds, often taking 10+ days for warehouse inspection. These measures frustrate good customers, as 68% of shoppers won't return after a refund takes over five days (NRF). Slow refunds also increase operational costs, becoming the second-leading cause of customer service inquiries (Narvar),

Presented in Riskified’s report, “Refunduary is here, How to prevent refund policy abuse, protect profitability, and optimize CX,” the data underscores why merchants feel pressured to apply these strict controls:

• The holiday season is particularly challenging: November and December orders generated almost a third of all claims for the year, with over half carrying into January and overwhelming operational teams.
• High-value orders are another area of risk: Claim rates for orders over $2,000 are 2.5x higher than those under $100, and orders over $1,000 are 33% more likely to be abusive compared to the average. For high-spending VIP customers, additional scrutiny can mar the experience.
• "Item Not Received" (INR) claims are the most abused category, with a 25% higher likelihood of fraud vs “Missing Items” (MI) claims, as bad actors exploit liability gaps (consistent with Dark Web examples). However, the surge in ecommerce order deliveries means there are plenty of legitimate INR claims.
• Fraudsters exploit early-claim windows: Claims filed within the first seven days are over 20% more likely to be abusive vs. the average, creating tension for merchants who want to provide fast service to legitimate shoppers.

“Fraudulent claims don’t just hurt retailer profitability, they cost good customers time and patience,” said Jeff Otto, Chief Marketing Officer of Riskified. “When refunds are delayed for days or weeks, loyal shoppers become frustrated and may not come back. Riskified’s AI platform with Dynamic Returns helps merchants quickly distinguish between abusive and legitimate claims by evaluating customer behavior and identity history in real time. That precision allows trusted customers to receive fast resolutions while risky behavior is handled appropriately, protecting revenue while keeping good customers satisfied and loyal.”

To help retailer fraud and customer experience teams improve the customer claims process, Riskified has introduced Dynamic Returns, a new feature within Policy Protect. Dynamic Returns enables refund and exchange options to be tailored in real-time based on customer risk and eligibility. Rather than applying uniform rules or delaying refunds until warehouse inspection, merchants have the identity insights they need to assess customer eligibility and trigger the appropriate outcome, using real-time return intelligence that evaluates customer behavior and network-level signals.

As soon as a return claim is submitted, retailers can act immediately: Trusted shopper identities can receive an instant refund, replacement items can be shipped without waiting for the original, and some refunds can be issued without an item being returned at all. This dynamic approach allows merchants to handle clear cases of abuse, recognize trustworthy shoppers, and manage the gray area in between.

Using the new Dynamic Returns feature within Policy Protect, a retailer was able to instantly approve more than half of returns for loyal customers. Customers who received early refunds recorded a more than 20% increase in customer satisfaction score (CSAT) compared to those who went through traditional refund processes, with over 97% of approved early refunds going to customers who actually returned their purchased items. (Results reflect outcomes observed over a one-month period.)

As ecommerce continues to grow, retailers face increasing pressure to balance convenience with control. Policy abuse isn’t just a financial problem--it’s a customer experience challenge. Merchants using Riskified’s Policy Protect to detect and prevent abusive claims can ensure their shoppers receive the service they expect, maintain trust, and protect revenue, all at the same time, in real-time.

Learn more about Policy Protect and Dynamic Returns on Riskified.com.

Refunduary Report Methodology

The findings in “Refunduary is here” are based on Riskified’s analysis of 1M+ refund claims from three major retailers over the full 2024 calendar year. The dataset includes all claims received, with a focus on “Item Not Received” (INR) and “Missing Items” (MI). Riskified’s proprietary data science models flagged claims as highly suspicious based on a combination of high-risk statistical patterns and behavioral indicators, allowing for a reliable assessment of policy abuse. Read the report here.

About Riskified

Riskified (NYSE:RSKD) empowers businesses to unleash ecommerce growth by outsmarting risk. Many of the world’s biggest brands and publicly traded companies selling online rely on Riskified for guaranteed protection against chargebacks, to fight fraud and policy abuse at scale, and to improve customer retention. Developed and managed by the largest team of ecommerce risk analysts, data scientists, and researchers, Riskified’s AI-powered fraud and risk intelligence platform analyzes the individual behind each interaction to provide real-time decisions and robust identity-based insights. Learn more at riskified.com.

View source version on businesswire.com: https://www.businesswire.com/news/home/20260112619982/en/

Corporate Communications: Cristina Dinozo, Senior Director of Communications | press@riskified.com

Investor Relations: Chett Mandel, Head of Investor Relations | ir@riskified.com

Source: Riskified