Glassbox helped SoFi correct a website issue resulting in multi-million dollar savings

Case Studies 15. April 2021.

The Company

SoFi is a digital personal finance company that started in 2011 by developing a lending solution for students with lower loan rates in an attempt to address the mounting student debt problem in the US. Now with over 1,500 employees, the San Francisco-based company has branched out into mortgages, personal loans, investing and banking—all with a digital-first approach.

The Challenge

SoFi was not aware of a technical issue on their website causing an increase in incomplete loan applications and was blind to the magnitude of the phenomenon. In fact, the issue went completely under their radar until Glassbox started to capture and visualize all of their user sessions.

The Solution

With Glassbox’s tag-less deployment, SoFi was capturing 100% of sessions and all events and elements, including user activity and technical elements occurring behind the scenes.

Root-casue analysis sped by real-time anomaly detection and alerts

Glassbox’s anomaly detection engine, based on machine learning, alerted the SoFi team to 546 sessions in a one-week period in which users received an error message and never completed their loan application.

Measuring the revenue impact with Cashbox

SoFi leveraged the Cashbox dashboard in Glassbox to capture the value of the user input in the loan application form. The team was able to aggregate all loan applications and track the following insights:

  • Average loan value
  • Monthly loan value
  • Average monthly loan value
  • Geolocation
  • OS/Browser/ISP

By recognizing and understanding the full scope of the problem and true monetary impact of this technical error, the business team and the IT team flagged this issue as a top priority.

By sharing session replays of the affected sessions with the production support team, the root cause of error was identified and fixed within the same day.

The Main Results

If the glitch had continued to go unnoticed, the failed sessions would add up to over 28,000 abandoned applications annually. With an average success rate of 38.2%, it was estimated this error could have caused a loss of 10,845 loan deals. Based on a hypothetical monthly loan value of $900 per loan (the actual loan value has been changed for demonstration purposes), SoFi prevented a potential annual loss of over $9 million by fixing this error.


"Glassbox gives us the most accurate view of how our members experience our app and saves us countless time on squashing bugs, finding friction, and gives us confidence in how we plan our UX roadmap."

Software Development Lead at SoFi

About Glassbox

Glassbox is a customer experience solution that enables companies to better understand their customers’ demands and struggles across all digital platforms. This GDPR-compliant solution automatically records all customer sessions on the website or app and creates dashboards with company-specific KPIs. Glassbox’s 100% tagless recording, intuitive real-time reporting and AI-powered analysis is a proven concept across industries where digital comes first. Besides SoFi, their existing clients include Singapore Airlines, John Lewis, SuperDry, Tracfone, Admiral, Hong Kong Express, the second largest telecom provider in the US, and six out of ten of the largest banks in the world amongst other industry leaders.


Additional Information

Homepage Glassbox
Case Study Glassbox & SoFi

Contact us to learn more

Stay updated. Subscribe to the MMV newsletter