How River Island used Ravelin to transform and expand its fraud strategy

River Island is one of the UK’s leading fashion retailers, operating 300 stores and six websites.

Before the integration of Ravelin’s solutions, River Island’s fraud strategy was heavily reliant on rules and manual review. By focusing on human decisioning, they had been effective in keeping fraud rates low – but this was not scaleable or an effective use of resources.

Nick Kirby, River Island’s Fraud Operations Manager, expands on the situation:

“The impact of running a seven days-per-week, 16-hours-per-day fraud prevention operation was significant for a small team. Coverage was a constant stretch and it impacted morale and motivation. At a time when I could see that the industry was getting smarter with resource deployment, it felt like the team was in danger of getting bogged down.

Having shifted to an automated decisioning approach through Ravelin, River Island was able to:

• Reduce active rules from 300+ to to less than 40 (an 87% reduction!)

• Completely eliminated manual review

• ...all while maintaining a consistently low fraud rates

Speaking in 2023, Grant Shipway, Global Fraud Manager at River Island, describes the collaborative relationship between River Island and Ravelin over this period:

“Our Islanders and the Ravelinos worked in harmony. We met every week and, once the pandemic allowed it, we would frequently meet in person for workshops and to discuss the longer-term strategy.

This built trust on both sides as, for us, we want Ravelin to enjoy working with us as much as we enjoy working with them.”

He went on to say:

Two years on, “third-party fraud prevention has been mitigated thanks to the collaborative efforts of Ravelin and River Island, which means we can turn our attention elsewhere.” says Lacey Cole, Fraud Analyst at River Island.

As Grant Shipway highlighted in his 2023 Ravcon talk, “we need to look at the entire customer journey and not just over-index on the point of payment”. To this aim, Grant set up the Profit Erosion Committee – a cross-functional internal team that identifies drivers of profit erosion for River Island, and sets up projects to tackle them.

Adopting a machine-learning-first approach has “given the team a huge amount of ability to move into different areas, despite actually having a reduction in headcount. For me that’s customer fraud, particularly around refund abuse,” says Lacey.

Successfully thwarting refund abuse presents an opportunity to protect even more revenue for River Island.

But, while the opportunity size is great, the nature of the problem is complex:

“Customer fraud is a lot more dynamic. It’s a lot harder to pinpoint compared to third-party fraud. For instance you can have a great customer that just pivots one day particularly with the cost of living crisis and the changes driven by the pandemic,” says Lacey, going on to explain:

“Covid rightly made it easier for our great customers to buy, return, and refund – as a result, we have to unpick any bad customer behavior that comes from this new customer experience ecosystem.”

River Island has a two-pronged approach for tackling refund abuse:

Opportunistic fraudsters

The first is around behavior change. As Lacey highlights, previously good customers can seek to abuse return policies.

Therefore, River Island is evolving a “nudge” strategy. After identifying a segment of customers with irregular refund behaviors outside of the normal baseline, an email will be sent to this segment.

The long term vision is that these “nudge” emails will be automatically triggered in River Island’s CRM by calling the Ravelin API after a certain threshold of “acceptable” refund behavior is crossed. They will then receive tailored emails based on their refund item, refund reason, and recent refund history. The goal? To salvage the lifetime value of these consumers, discouraging them from seeking abusive refunds but still enabling them to continue to shop on the website

Professional fraudsters

However, for refund abusers who are clearly fraudsters, the goal is to prevent those individuals from transacting with River Island again. “They disrupt our ability to serve our thousands of great genuine customers and this isn’t something we can stand by and allow” says Lacey.

Therefore, the system will block the individuals that are identified as outright fraudsters with an aim to never give them the opportunity to attempt fraud against River Island again – as well as use link analysis and machine learning to ensure that they will not be able to create new accounts.

Ravelin collects rich data on each customer, presenting River Island a holistic customer view in one place.

Lacey and team began by analyzing existing customer refund data using Ravelin’s Data Warehouse, which houses all their customer data. “Ravelin actually started the SQL journey for our team. Until last year, none of us had any experience with SQL. It’s great for a team of analysts to be able to pull and query the data ourselves. It’s awesome that Ravelin supported us on this journey of self-learning,” notes Lacey.

Through analyzing customer history, returns behavior and reason code, predicted lifetime value and so on, all captured in Ravelin’s solutions, River Island was able to identify “key baselines for unusual refund behavior and compartmentalizing our customer base to understand who may be ‘bad’ customers or someone we should be keeping an eye on,” said Lacey.

Using Ravelin as a source of truth, River Island works cross-functionally with its distribution center and customer service team to tag customers who exhibit suspicious refund behavior in their Ravelin to continue to build that profile of the customer’s behavior.

Part of this is identifying connections between known and potential refund abusers using link analysis. “Using the Connect Graph Network to identify connections between refund abusers and other fraudsters has been fantastic.

It's probably the best fraud tool I’ve seen that shows the connections in such a visual way. I’ve used other fraud tools and you really have to dig for the information. But with Ravelin, it shows so clearly, and so cleanly. It’s been really useful,” Lacey says.

• Ravelin insights: Policy abuse
• Webinar: River Island's journey from fraud rules to machine learning
• Are you spending too much time on manual review?
• Refund abuse: Why is it on the rise and how do we stop it?
• Brazen consumer attitudes revealed in Ravelin’s Consumer Fraud Survey
• Are merchants facing a criminal customer crisis?
• On-demand webinar: How to solve refund abuse