How a global domain registrar freed up time and beat fraud

Technology Category

Analytics & Modeling - Machine Learning

Applicable Industries

Telecommunications

Applicable Functions

Sales & Marketing

Use Cases

Fraud Detection

Services

Data Science Services

The Challenge

iwantmyname, a global domain name registrar, was facing a significant challenge with fraud. The company was losing 2% of its revenue to fraudulent activities, which was unsustainable given the competitive nature of the industry. The process of detecting fraud was entirely manual, with two of the co-founders checking every single transaction for suspicious signals. This was not only time-consuming but also led to the company blocking all users from countries with high levels of fraud, negatively impacting their business. The company was missing out on revenue from legitimate customers in these countries and existing customers traveling in these countries faced extra hassle with their accounts. The team was spending as much as 30% of their time managing fraud, time that could have been better spent on growing the business and improving customer experience.

About The Customer

iwantmyname is a global domain name registrar based in New Zealand with around 100,000 global users. The company offers a free DNS service, robust customer support, and an easy one-click import that automatically loads required DNS records for all of a customer’s favorite hosted web services. iwantmyname prides itself on being the ethical registrar of choice, actively working in the community, supporting startups, and giving back a portion of every registration fee for charitable donations towards environmental sustainability projects and disaster relief. The company's mission is to provide an excellent customer experience by making the process of purchasing a domain name simple, clear, and transparent.

The Solution

iwantmyname implemented Sift, an automated solution that uses machine learning technology to detect and prevent fraud. The integration was completed in a single day by two developers and began catching fraud immediately, dramatically reducing both fraud attacks and nuisance transactions from credit card testers. The team incorporated the Sift Console into their daily operations after a couple of days of training. Within weeks, they experienced a 75% reduction in fraud on their site by using the Sift Console to make their decisions. The company was able to lift their country blacklists and begin accepting orders based on Sift’s sophisticated machine learning-based risk scores, leading to higher accuracy, a better customer experience, and more revenue.

Operational Impact

The company was able to reduce fraud on their site by 75%.
The team was able to incorporate the Sift Console into their daily operations, leading to more efficient fraud detection and prevention.
The company was able to lift their country blacklists and begin accepting orders based on Sift’s sophisticated machine learning-based risk scores.

Quantitative Benefit