In-store transactions are often perceived as lower risk compared to online card-not-present purchases. EMV chip cards, card-present authentication, and protections from card networks and processors create a sense of security that can make additional fraud tools appear unnecessary, expensive, or even disruptive to the checkout experience.

What’s more, because effective fraud protection stops fraudulent transactions before they happen, merchants rarely see the losses it prevents. This can create a dangerous illusion: fraud appears negligible after it is successfully suppressed over time.

But when two retailers put this assumption to the test by pausing in-store fraud reviews, the results were swift and costly.

The scenario: in-store fraud protection at leased registers

Many department-style retailers operate using a leased register model. In this setup, individual brands run their own mini-stores within a larger retail location. These brands staff their own associates and process transactions through dedicated registers, even though they operate under the umbrella of the larger retailer.

This model is particularly common in luxury retail, where brands prioritize highly trained staff and a consistent, on-brand customer experience.

As part of its leasing agreement, one high-end department store had historically included Riskified’s in-store fraud protection and chargeback guarantee for its partner brands, allowing them to operate securely in their stores.

As it does for online transactions, Riskified’s chargeback guarantee means that Riskified assumes the fraud risk and covers the cost if a transaction later results in a fraud-related chargeback. Merchants can confidently serve customers in-store knowing Riskified will reimburse eligible chargebacks and handle the dispute process. 

A costly experiment: pausing fraud protection in-store

In 2024, the department store made a strategic change to its leasing model. Rather than covering the cost of in-store fraud protection, it shifted that responsibility to its partner brands. Each vendor could choose whether to continue using Riskified’s fraud protection at their own expense or opt out entirely and process transactions without fraud reviews.

To some brands, opting out appeared reasonable. Thanks to Riskified, approval rates were high, declines were rare, and visible fraud seemed minimal. But based on historical data and the fraud it was blocking behind the scenes, the Riskified team understood the likely outcome of pausing fraud decisioning.

To make the risk tangible, Riskified created individualized forecasts for each brand. These projections estimated the chargeback costs vendors could expect if fraud protection were removed, using historical approval rates, previously declined transactions, and known fraud patterns to translate risk into bottom-line financial impact.

Presented with these projections, most of the store’s partner brands chose to retain fraud protection; but two major luxury brands — Brand A and Brand B — did not. They decided instead to pause fraud protection and observe the impact for themselves.

The results: chargebacks surge without protection

Within months, both brands saw their chargeback costs spike, close to Riskified’s projections. As fraud increased, operational strain followed. Support tickets surged, internal teams were pulled into dispute management, and the financial impact became impossible to ignore:

  • $1.3M+ in chargeback losses over six months for Brand A
  • $800K+  in chargeback losses over six months for Brand B

Reinstating fraud protection: losses drop to near zero

Faced with mounting losses and a clear demonstration of underlying fraud risk, both brands eventually reinstated Riskified’s AI-powered fraud reviews and chargeback guarantee. Once protection was restored, fraud-related chargebacks quickly dropped to near zero.

The test proved expensive, but it delivered an unambiguous lesson: even at the register, fraud protection provides measurable and ongoing value. In this case, the losses incurred proved that Riskified was delivering significant ROI when considering chargeback costs alone, exclusive of operational costs and long-term harm to the brands’ customer experience:

  • Brand A – Riskified ROI: 1600%
  • Brand B – Riskified ROI: 600%

Chargeback guarantee pays for itself

Fraudsters continuously test retailers for weaknesses, and when a gap appears, they exploit it quickly and aggressively.

This case demonstrates that effective fraud protection does more than prevent losses. It pays for itself many times over by reducing chargeback costs, preserving a seamless experience for legitimate customers, and protecting long-term customer loyalty and lifetime value.

  • Riskified machine-learning models protect revenue by approving good orders and rejecting fraud in real time — without customer friction
  • Riskified guarantees approval rates and is financially accountable for chargebacks
  • Riskified converts payment failures and false declines into sales, driving additional revenue
  • Fewer declined orders build deeper customer loyalty

Learn more about Riskified’s chargeback guarantee and how it protects retailers online and in-store.