Challenge

Cleo is an AI assistant that uses a playful sense of humour to help users save, budget, and build a healthy relationship with their money. To access additional services such as interest-free cash advances that users can repay later when they can afford it and credit score building, users can subscribe to its Cleo Plus or Cleo Builder subscription tiers.

Previously, Cleo processed both subscriptions and cash advances through a legacy payment provider. Both types of payments were being declined at high rates, which cost Cleo revenue and forced the company to pay fees associated with the failed transactions.

“If our payment processor is declining transactions and payments keep failing, it’s bad for our business, because we lose money and our users can no longer access the services they need,” said Callan Carvey, global head of operations at Cleo. “We needed a reliable payment system to increase our rate of successful transactions and enable us to take money successfully.”

Cleo was also limited in its ability to prevent fraudulent transactions, lacking enough data to spot suspicious patterns of activity as they occurred. This meant its team spent considerable time identifying fraud after the fact in order to update its fraud model and prevent additional issues in the future. Cleo wanted to find a way to stop fraud before it happened, particularly to mitigate chargeback fraud while saving its in-house teams from time-consuming manual dispute tasks.

Solution

To improve its rate of successful transactions, Cleo implemented Stripe Payments for its subscription payments and its cash advance product. Stripe’s direct integrations with card networks and its Adaptive Acceptance feature, which uses machine learning models to optimise and retry declined payments in real time, Cleo began to reduce the likelihood of customer payments being declined.

Cleo also integrated Stripe Radar for Fraud Teams to identify and prevent fraud. Trained on hundreds of billions of data points, Radar for Fraud Teams uses early fraud warnings to alert Cleo to suspicious patterns of activity and automatically blocks high-risk payments. Cleo also uses Radar for Fraud Teams to create custom block lists to prevent attacks from users with specific emails, card numbers, and IP addresses. With its customisable protection, Radar for Fraud Teams allows Cleo to write fraud rules tailored to its business needs and risk tolerance.

For suspicious chargebacks, Stripe’s integration with Chargehound – a comprehensive chargeback solution that automates chargeback disputes – saves Cleo from having to contest chargebacks itself. Stripe passes details of contested payments via an API to Chargehound, which then automatically handles the dispute process with the card issuer. As a result, Cleo is able to recoup lost revenue from fraudulent chargebacks, without having to spend time and resources disputing chargebacks.

Results

1.5% increase in cash advance repayments

With Payments, Cleo has seen a 1.5% increase in successful cash advance repayments. This means Cleo receives more money from successful payments, while reducing failed transaction costs and increasing overall revenue.

“As a company, we talk about everything in terms of lifetime value, with a dollar value attributed to every user,” Carvey said. “Leveraging Stripe for our cash advance business reduces our failure costs by increasing our success rates, resulting in a positive impact on our user lifetime value.”

23% lower chargeback rate on cash advances

Radar for Fraud Teams helped Cleo achieve a 23% reduction in the rate of fraudulent chargebacks on its cash advance service. This saves Cleo considerable lost revenue, while also saving Cleo’s fraud team significant time and resources manually aggregating and analysing data to fight fraud.

“Radar for Fraud Teams enables us to keep pace with fraudsters by leveraging inputs and trends to inform our own fraud rules internally, meaning we can spot fraud in near real time and prevent it from happening,” Carvey said. “If it wasn’t for the Radar Dashboard, the early fraud warnings, the customisation, and more, our fraud team would need to be triple the size it is today, if not bigger.”

$15,000 saved per month with automated chargeback disputes

By automating chargeback disputes through Stripe’s integration with Chargehound, Cleo has been able to automatically contest fraudulent chargebacks, resulting in savings of $15,000 a month in recovered chargebacks and saved dispute fees. This amounts to over $200,000 saved since implementation, without requiring Cleo’s fraud team to spend time disputing the reversed payments manually.