Reinventing the wheel with Metromile
Problem
Filing an insurance claim after a car accident is often a frustrating and time-consuming process even for simple cases like roadside assistance or glass repair. Internally, insurance companies faced their own struggles—claims adjusters were costly and limited, leading to bottlenecks that slowed down response times. How might we create a more seamless and automated experience for customers while reducing the workload for claims adjusters?
Research
We conducted a survey with 66 Metromile users. From the research, we found that most of the pain points came from the communication of the status of a repair, rental and the payout expectations and timeframe. We were able to plot out the different phases diving into the happy and not so happy paths.
Ideation
While we were in the ideation phase, we constrained the project to focus on automating single party incidents which accounted for anywhere between 10-20% of total claims. We broke down the phases for selecting a repair shop and a rental car or lyft ride if needed. We explored different ways of communicating these areas. We also explored translating our AI model as an in-product experience.
Results
To ensure accuracy and prevent fraudulent behavior, we developed an intelligent claims model leveraging Metromile’s plug-in tracker and driving behavior data. This model assessed claims based on accident circumstances, flagged potential fraud, and automatically approved legitimate cases within minutes. This not only streamlined the experience for honest customers but also safeguarded the company against unnecessary payouts.
Contributions:
Strategy | Interaction | Visuals | Prototyping | IA