Improving Gig Worker Check-ins: Reducing Fraud and Simplifying the Process

Anthor’s gig worker check-in system faced critical challenges that threatened the platform’s efficiency, security, and user experience. Workers frequently encountered technical glitches, time-consuming check-ins, and user errors, while the system's vulnerability to identity fraud compromised trust and reliability.

To address these pressing issues, I led a redesign that simplified the clock-in process, introduced a seamless one-click check-in feature, and resolved underlying technical glitches. This overhaul not only improved the user experience but also provided real-time, data-driven insights that strengthened operational efficiency and security.

A key component of the redesign was the introduction of selfie and GPS verification, adding robust layers of security. Selfie verification allowed real-time identity confirmation, while GPS verification ensured workers’ physical presence at the required job location. This combination reduced the risk of fraud and improved accountability.

The redesigned system delivered a smoother, more reliable experience for gig workers, while significantly enhancing Anthor’s operational security and trustworthiness.

Before the redesign, users faced a badly formatted error message whenever they tried to clock in at the wrong time or location.

My contributions

Disclaimer: This is a small part of the complete study done for the project.

A key component of this project was conducting in-depth interviews with all the teams that interacted directly—and in some cases, indirectly—with the check-in feature. These conversations uncovered valuable insights into the bottlenecks and operational issues caused by the existing system. Understanding how these challenges impacted not just users, but internal team workflows and scalability, was essential in shaping a solution that could address both user experience and back-end efficiency.

Next, I deconstructed the user flow and requirements. It became evident that the process was overly complicated, requiring at least six steps for a successful check-in but having 16 different failure points. Users had three different paths they could take, each with its own obstacles, making the process confusing and frustrating.

The root of many issues was the reliance on two main requirements for success: the store needed to have a valid QR code displayed somewhere in the building, and the worker had to be within the correct GPS coordinates. During the interviews, I discovered that the QR code requirement was a significant bottleneck for our B2B onboarding team. As we scaled, it became harder to provide each store with its unique QR code in time, since our team either had to visit the stores personally or ship the codes via a welcome kit. This logistical issue led to a growing number of gigs being requested without the necessary QR code in place, further complicating the check-in process for workers.

before

(click to zoom in)

In the previous user flow, if errors with the QR code and location occurred simultaneously, users were directed to a live chat with the support team. This manual process required the support team to verify the user's location by asking for a Google Maps screenshot and a picture of the store front. Only after these verifications were completed could the gig be started, leading to significant delays and frustration for users while creating a time-consuming workload for the support team.

after

The redesigned user flow eliminated unnecessary alternative paths, streamlining the process for a smoother experience. One of the key changes was providing users with a simple way to report location errors. In the new system, the app automatically sent users' current coordinates to our back-end, allowing us to confirm whether they were at the correct location. While our back-office team had been validating coordinates manually, this update provided quantifiable data on how often each store's location was flagged. This made it much easier to prioritize and fix problem areas, significantly improving operational efficiency.

Another critical improvement was removing the “failure” result that directed users to the support team. By automating validations within the app, errors were shifted to a post-gig verification process. This not only gave our security team valuable data to track fraud but also ensured that users could proceed with their gigs without delays or waiting on support, leading to a more seamless and efficient experience.

After the redesign, users had a lot more information about what was expected of them: they had access to a map indicating the check-in location and GPS range.

Results achieved

  • Reduced identity fraud cases by 80%, including shutting down a fraudulent operation run by users on the platform.

  • Reduced support tickets related to errors by 95%, significantly decreasing the workload for the support team.

  • Reduced the number of gigs requiring technical support from an average of 400 daily to 20, improving operational efficiency and allowing gig workers to start their shifts without delays.

Graph shows total events of gigs completed (activity-finish) and reported issues (activity-execution-problem) over a year. Within 3 days of the launch we saw the reported issues plummet and stabilize while the gigs completed went on a steady rise over the next couple of months.

Next
Next

Driving Accountability and Improving Workforce Reliability in an On-Demand Staffing Marketplace