Kogan lifts approval rates above 98% with Riskified

Kogan has expanded its use of Riskified's fraud and risk tools across its online retail operations, helping lift approval rates above 98%.

The Australian eCommerce group also identified AUD $1.5 million in annual savings through better detection of policy abuse and promotion misuse, while reducing manual work tied to chargebacks and order reviews.

Kogan serves more than 3.5 million customers across Australia and New Zealand and operates entirely online. As a result, every purchase is a card-not-present transaction, a category with higher fraud risk than in-store payments.

The issue extended beyond fraudulent card use. Kogan also faced losses from serial policy abusers, promotion misuse and subscription-related chargebacks, which were hurting margins and adding operational pressure.

That created a difficult balance between blocking suspicious transactions and avoiding false positives that can turn away genuine customers. Kogan wanted to strengthen fraud controls without making checkout harder for shoppers.

"We wanted a solution that could truly perform, without impacting customer experience," said Rohan Dhaygude, Deputy CFO, Kogan.com.

Kogan adopted Riskified's Chargeback Guarantee model alongside identity analysis and automated dispute management tools. The work focused on improving fraud prevention, increasing approval coverage and building a clearer picture of customer behaviour across the shopping journey.

Abuse detection

Kogan used Riskified's Identity Engine and Identity Explore products to track repeat policy abusers, identify promotion misuse and assess customer-level profitability. Device data and behavioural signals helped distinguish fraudulent activity from legitimate shopping more accurately.

Machine learning models trained on Kogan transaction data were also used to refine decisions on higher-value purchases and first-time customers. That segment is often harder for online retailers to assess because limited purchase history can make legitimate shoppers appear riskier than returning buyers.

Chargeback handling was another focus. By combining automated dispute management with a guarantee model, Kogan lowered manual workload, improved dispute outcomes and kept chargeback rates below AusPayNet thresholds.

The result was stronger approval performance in high-value and higher-risk categories, where retailers often face the greatest tension between fraud prevention and conversion.

Approval rates are a closely watched metric in eCommerce because unnecessary declines can directly reduce sales. For large online retailers, even small shifts can have a noticeable effect on revenue and customer retention.

Customer intent

The work also improved Kogan's visibility into customer behaviour, giving teams more confidence to approve legitimate orders. That matters during busy retail periods, when transaction volumes rise sharply, and merchants can struggle to maintain review standards without adding staff.

"Riskified stands behind its decisions, creating a strong alignment of incentives and giving us confidence to approve more legitimate orders while reducing fraud exposure," said Dhaygude.

Kogan's online retail and services portfolio includes Kogan.com, Dick Smith, Mighty Ape, Matt Blatt and Brosa, alongside mobile, internet, insurance, money, energy and travel offers. That breadth means fraud and policy abuse can emerge across several transaction types rather than in a single channel.

Riskified, which is listed in New York, sells fraud prevention and chargeback protection tools to online merchants. Its model centres on assessing the identity behind a transaction and making a real-time decision on whether an order should be approved.

For retailers, the commercial case for such systems extends beyond fraud losses alone. Reducing manual reviews, handling disputes more efficiently and limiting policy abuse can all improve margins, particularly for businesses with high transaction volumes and thin retail economics.

The Kogan results point to a broader challenge for eCommerce groups in Australia and New Zealand as they try to protect against fraud without disrupting customer experience. Retailers are increasingly under pressure to identify not only stolen payment activity, but also less visible forms of abuse that can erode profitability over time.

"What stood out with Kogan is how fast they moved from just flagging risk to actually understanding their customers' intent," said Aviram Ganor, GM EMEA & APAC, Riskified. "That's what lets them say yes to more good orders instead of playing it safe, ultimately creating a superior experience for Kogan customers."