Challenge

Ben & Frank is the most popular direct-to-consumer eyewear brand in Latin America. With its previous payments provider, Ben & Frank had limited fraud control and payments reporting, making it impossible to continuously optimise authorisation performance. This is because Mexico’s market has some unique challenges:
- Latin America is well known for being a highly customised and localised payment ecosystem, making it difficult to be successful without customised user payments systems or close relationships with major issuers.
- Buyer fraud rates are high by global standards, making it challenging to define an effective approach to suspicious charges and dispute management.
- There's limited support for e-commerce features in Mexico. For example, it's difficult for businesses to show issuers that repeat customers with saved card credentials are less risky than first-time customers.

As a result, Ben & Frank had low conversion rates, which created a poor customer experience for those who received an incorrect decline.

Solution

Ben & Frank compared Stripe to its existing payments processor by measuring authorisation rates across card networks. John Campbell, Operations Director at Ben & Frank, said, "In evaluating a new payments processor, we were looking for three measurable goals: acceptance rate, fraud rate and NPS of users’ experience with our customer service.”

Ben & Frank was able to integrate and launch Stripe in just two weeks to begin testing. Leonardo Alonso, Software Engineer at Ben & Frank, said, “Stripe has an amazing team ready to support our needs. It has one of the best development workflows, and the documentation is very easy to use with extensive client libraries and copy-and-paste examples.”

After the three-month trial, Ben & Frank determined Stripe was the better payments solution to improve conversion and power its continued growth, thanks to its responsive support and modern suite of solutions featuring payments optimisation and fraud detection.

Results

Ben & Frank saw a 10% uplift in conversion in comparison to its previous processor in just three months of using Stripe in Mexico. The uplift Ben & Frank experienced was made possible by:

Stripe’s close relationship with regional processors and issuers
Stripe hires dozens of local payments experts and developers to build strong partnerships across Mexico’s payments ecosystem, inclusive of the local switches like Prosa and eGlobal, Banco de México and CNBV, and the large issuers like BBVA and Santander. To establish trust with local issuers, Stripe meets frequently with the issuers to make iterative changes to authorisation and fraud logic, helping businesses capture revenue that would otherwise be lost.

Reduction in fraudulent activity and resolution time
Because Radar is built on top of Stripe Payments, Ben & Frank no longer has to manually label fraudulent transactions. Instead, the business relies on Stripe’s machine learning to evaluate each payment risk level in real time. The machine learning model continuously learns from new customer purchase patterns, while Ben & Frank's fraud team can use manual rule overrides and custom fraud rules that allow for a localised strategy to proactively manage fraud and chargeback rates in Mexico.

With Latin America fraud rates at 5x that of the United States and EMEA, Radar helps Ben & Frank distinguish fraudsters from customers, reducing fraud at a greater rate. Lourdes García, Product Manager at Ben & Frank, said, “By leveraging Stripe’s machine learning and custom fraud rules, we've seen a decrease in our fraud rates and chargeback rates. We’ve even found our customer satisfaction has improved because our team settles customer disputes faster now that they spend less time managing a manual process.”

Authorisation rates optimised by machine learning
Banks have inconsistent decision rules on which transactions get accepted. The only way to improve acceptance is by reverse-engineering the rules across the Stripe network using machine learning. To do this, Stripe built Adaptive Acceptance, a machine-learning technology that optimises authorisation messages on behalf of its customers. Stripe’s data scientists and engineers continuously improve these machine learning models to help businesses accept as many legitimate transactions and revenue as possible.