Accurate Data Tracking on Web to Boost User Engagement

Problem

Loco's web platform faced a significant challenge as it solely transmitted data to the product analytics client, Amplitude, without sending crucial information to the backend. This limitation was particularly pronounced because Amplitude only allowed data querying at the event level and lacked the capability to provide user-level insights at scale. The complexity escalated when dealing with critical metrics such as tracking users who viewed live streams or Video On Demand (VOD) content on the web. Unlike the app, web-based viewership introduced diverse scenarios, including content running in the background or simultaneous watching across different windows/tabs. This lack of synchronization between the web and backend systems posed a substantial hurdle in accurately tracking user engagement, especially in cases where minimum watch durations were tied to rewards and reported as crucial business metrics monthly.

Solution

To overcome the challenge of untracked user engagement on the web platform, I undertook a comprehensive approach. Recognizing the diverse ways users consumed content on the web, I meticulously identified all potential use cases and edge cases. Working collaboratively with web developers and quality assurance (QA) teams, I devised a solution that involved triggering key events related to watch duration and live stream/VOD visits at precise moments. The solution aimed at addressing each identified use case, ensuring accurate data tracking on our backend service.

Through extensive testing, we systematically eliminated all identified edge cases, refining the event triggering mechanisms to capture nuanced user interactions accurately. This process involved resolving issues related to background content playback and parallel watching across multiple windows or tabs. The solution sought to enhance data accuracy, particularly for critical business metrics tied to rewards and monthly reporting. This solution also positioned us better to handle streamer payouts based on web watching, once it's integrated.

Impact

• 15-20% increase in average watch duration per logged-in user on the web

• Improved personalization of the web platform for better user experience