Tradesy is a popular ecommerce platform for buying and selling luxury clothing and accessories. The peer-to-peer marketplace is on a mission to bring trust and fairness to fashion, making great style affordable and accessible to everyone.
According to Tradesy, closets shouldn't lay littered with unwanted wear: a once-worn wedding dress, beautiful shoes that pinch your feet, or pants that no longer fit. Instead they should be sold to people who will love them. Millions of people who use the platform agree. Tradesy is quickly becoming the go-to exchange for luxury goods.
Tradesy's data engineering team had implemented Google Analytics to understand shopper behavior, hoping to turn browsers into buyers with great app design. Google Analytics helped give insight into customer behavior, but it only illuminated simple aggregated conversion metrics, like for instance, if a customer landed anywhere on the site or app and eventually purchased something. In addition, the team couldn't tie funnel data to actual users since Google Analytics anonymizes every interaction.
In order to improve conversion rates and learn more about their customers, the team needed more information. Why were customers bouncing? What made them stay? Where exactly in the checkout process did they abandon the chase?
Tradesy's CTO considered building a solution in house to consolidate data across their apps, websites, and transactions. However, building a data collection library, ETL pipeline, and analytics integrations that were reliable, scalable, and worked across devices would have been a massive undertaking — requiring three people for six months to be exact.
We played around and built a couple of prototypes for our own data pipeline, but we quickly realized that it would take three engineers six months to stabilize a platform, and even then it would not be as feature-rich as Segment.
Tradesy chose Segment as their Customer Data Platform. They use Segment’s collection libraries for the web and SDKs for Android and iOS to capture valuable user data in a granular fashion. Segment reliably delivers the data to their end tools, helping them understand how people come to their platform, what exactly they do, and whether or not they buy or come back.
The best thing about Segment is that it just works. When a customer triggers an event in any of our mobile apps or on our website, the data shows up in Segment and then our connected destinations without fail.
Once they had instrumented Segment, the team had the option to send their data to any of the 200+ tools on the Segment platform with a few clicks. Tradesy chose to load their data into Amazon S3 via Segment’s integration.
Amazon S3 offers a scalable storage platform for raw data that the team uses to power exploratory analysis on the checkout funnel and other user flows. They also use the data to feed machine learning algorithms that provide product recommendations to each customer.
Since implementing Segment, Tradesy has been able to gain a deeper understanding of the customer journey. They now know exactly when, why, and how people get stuck before buying or listing a fashion item.
With Segment, we’ve been able to ramp up the number of data points we’re collecting from 5 to 100. This is huge for us because we’ve significantly reduced our time to insight. Before Segment it would take weeks, or months even, to implement any new event. Now it’s a few minutes.
Tradesy is using the new data insights to focus their checkout redesign efforts on eliminating the biggest friction points. For instance, they noticed better conversions in their native mobile app compared to mobile web, and used Segment to identify mobile web browsers and drive them to install the app. To make this analysis easier, they used Segment to consolidate their analytics platforms. They transitioned from a legacy solution to using Interana and Tableau, two business intelligence tools that leverage raw Segment events.
With the data pipeline project and maintenance off their plate, Tradesy’s data engineers have been able to work on more interesting challenges, like filtering out bot traffic from their analytics and making their recommendation engine operate in real-time.
"Segment has been a pleasure to work with on both the product and the support side," said Michael Viamari, data engineer at Tradesy. "Implementing it was easy and straightforward, and it has saved us a significant amount of engineering time and effort that we can now devote to leveraging our data to improve our product."