Customer complaints are hard to hear, but they give great insight into how to improve a product or service. For most fintech companies, data on customer complaints is unstructured. How can you improve your product or service if you can’t hear what your customers are saying?
A San Francisco based fintech giant has 400M+ users and started to see churn among their members. Although they had support tickets, the data was unstructured and they weren’t able to identify the specific pain points their members were facing. We partnered with data tooling company Tasq.ai, to create a data labeling solution that identified pain points for this client.
For all of our projects at Yenda, we source individuals who have industry-aligned degrees through internal tests and interviews, create a client specific Google classroom to train selected teams on the project, and establish a quality assurance process to guarantee high-quality output.
The goal of this project was to label customer complaint data to give the client actionable insights. Project details:
Data must be labeled to have actionable insights that improve customer retention.
Tasq.ai has a world class data annotation tool where they have a crowd of 100M+ individuals available to label projects that come their way. However, with an unmanaged crowd of that size, there are costs. Namely, visibility to who is doing the labeling, no communication with labelers, and limited industry-specific knowledge. For complex projects in need of industry specific talent, Tasq uses a Yenda team.
The project above required individuals familiar with finance terms and concepts, such as account limitation, transfer funds, card declines, and many more. Furthermore, Tasq has full visibility of the Yenda team, work completed, and a direct communication channel. The pairing of college-educated finance students with Tasq’s labeling tool resulted in accurate labeled data that provided actionable insights to address the key pain points for the client’s 400M+ users.
