- Technology company that makes marketing analytics software for online marketers
- Community of 600,000 marketers worldwide
- Faced both credit card and account fraud
- Existing solution couldn’t keep up with agile fraudsters
- Easy integration with API and JS snippets
- Automation based on Sift Scores
- Quickly and easily auto-ban bad users
- API logs and Network Visualizations allow for preemptive fraud management
Empowering businesses with smart tools
Moz makes marketing analytics software that helps businesses grow their SEO, content, social, and local marketing. With an expansive online community of marketers, Moz empowers users to grow their digital visibility through a suite of industry-leading research and analytics tools.
Founded in 2004 by Rand Fishkin and Gillian Muessig, Moz has more than 40,000 customers and 600,000 members of their worldwide community of marketers. Although most of their customers are U.S.-based, Moz has a growing presence globally with online marketers managing English-language websites.
Keeping the online community safe
As any growing business knows, fraud is always lurking among the good users. The fraud that Moz faces can typically be categorized as either payments fraud or account fraud. In the case of payments fraud, malicious users create new paid accounts with stolen credit card credentials with the goal of credit card testing to see which numbers are still viable. With any new paid account, Moz places a $1 pre-auth on the card number, but doesn’t actually charge the card. Nonetheless, these fraudsters can end up testing hundreds of credit cards on Moz’s site, which pollutes the order management system and the online community.
For Moz, account fraud is the more pervasive and damaging problem. Spammy users can sign up to join the community for free, but then use that access to try to phish for sales or try to game the Moz platform by creating hundreds of fake accounts and scamming Moz’s API keys. While the Moz team had a fraud management solution in place to combat credit card fraud, it couldn’t address the issue of spam accounts. Using simple rules – like “Does the zip code match the credit card information?” and “What is the country of origin for this card number?” – this solution couldn’t keep up with Moz’s expanding suite of offerings. As the user base grew and the team had to spend more and more time tweaking rules and examining suspicious users, the company sought a smarter solution.
Quick integration for easy automation
Managing this growing threat falls under Devin Ellis’ purview within the Engineering Team. Faced with the need to not just address existing fraud but preempt payments and spam accounts fraud, Devin decided to integrate with Sift. The integration process was simple, since the team already had an API-based solution in place. After swapping APIs, sending historical data points across, and adding Sift’s JS snippet, Devin’s team just needed to begin sending whatever events they deemed relevant and label the good and bad users. Within two weeks, Sift began to show accurate results.
The Moz team is now able to auto-flag suspicious users and auto-block malicious ones based on Sift Score. The Sift Score API allows the team to automate communication with flagged accounts, streamlining the Customer Support team’s response time and creating an overall better experience for all members of Moz’s Community.
“It’s a nice, intuitive UI that shows us a lot of data in one place.”
Set it and forget it
With a seamless integration with the Billing and Help team systems, Sift has – in Devin’s words – become a “set-it-and-forget-it” kind of solution. Moz leverages the features that showcase user timeline and page views, illustrating a suspicious user’s behavior and trends. The API logs are especially powerful, allowing the teams to see what’s happening across the board – and to effortlessly identify and ban repeat bad users, as well as visualize linked accounts.
The Sift solution saves Moz time and resources, as well as preserving its online reputation. Instead of wasting time hunting down fraudsters, the company can continue to grow its good user base and offer high-quality services to businesses worldwide.
“When we have repeat bad users, we can out-right block them based on Sift Score. Sift learns and catches them, every time they come back.”