A marketplaces and communities case study

How Poshmark reduced spam and created a secure online community

 

Poshmark is the largest social commerce marketplace for fashion, where anyone can buy, sell and share their personal style. With over 4 million Seller Stylists and millions of shoppers, Poshmark strives to keep an open and trusted environment for their community.

60–70%

Reduction in spam content

More frictionless environments for valued users

Fewer chargebacks

Told from the perspective of

Robbie Fritts,Director, Fraud and Payments
“Our partnership with Sift Science has really allowed us to focus on growth and enable our community to have a secure and trusted environment where they can share their personal style.”

Overview

Largest fashion marketplace in the U.S.

Poshmark brings together fashion enthusiasts on a platform that allows them to express themselves and share their love of style by buying and selling clothing. Their massive and ever-growing catalog features 75 million items from 5,000 brands. The ease of selling on Poshmark has been integral in growing their community of Seller Stylists to over 4 million, but with an increase of legitimate users comes an increase in fraudsters as well. Poshmark turned to Sift Science to mitigate spam content and fraudulent transactions in order to foster a trusted environment.

Challenge

Eliminating abusive content to maintain user trust

Community and trust are key to Poshmark, and are the foundation on which the platform is built. Phony listings, fraudulent buyers, and spam content erode user trust and discourage continued business on the app, while exposing users to dangerous actors. Poshmark needed a way to maintain a frictionless environment for trusted users, so they chose Sift Science’s Content Abuse Prevention product to get the job done.

“Trust is integral to building a strong community. We want to ensure people can express themselves in a trusted environment – whether that’s sharing personal style, liking listings, or buying. At the core, trust sets the foundation for a positive experience.”

Solution

Machine learning models that adapted quickly

The Content Abuse Prevention product allows Poshmark to take a forward-thinking approach to fraud management. Within a week, Poshmark trained Sift Science’s algorithms on live data and saw actionable results. Their fraud team became more agile, reacting faster to stop fraudulent activity, and if the team began to notice new trends they were able to leverage their decisioning and workflow tools to make tweaks in real time.

“Sift has an extremely high accuracy rate, which allows us to make sure we provide the least amount of friction for our beloved community.”

Results

Spam content greatly reduced, fostering a safe online community

Poshmark’s fraud team has a lot less work on their hands, thanks to their newly automated workflows, which has allowed the company to focus more on growth and less on fraud. They’ve seen a 60-70% reduction in spam content, which is crucial in maintaining the safe, trusted environment Poshmark buyers and sellers have come to expect from the platform. With Sift Science, Poshmark can put their efforts into supporting their community and feel good knowing their users – the core and heart of the company – are safer.

A network of trust

Online businesses face a variety of fraud and abuse threats. With over 12,000 sites on our platform, we collect, analyze, and learn from 100s of millions of events every day. Let’s fight fraud together.