In Zoomcar, Customers report discrepancies between the charges quoted during booking and the final amount billed to their accounts. As a product manager, how would you debug?

Introduction

At Zoomcar, the discrepancy between quoted charges and final billing amounts is a critical issue that directly impacts customer trust and satisfaction. I'll approach this problem systematically, focusing on identifying the root cause, validating hypotheses, and developing both short-term fixes and long-term solutions.

Step 1

Clarifying Questions (3 minutes)

• Looking at the booking process, I'm wondering about the frequency of these discrepancies. What percentage of bookings are affected by this issue?

Why it matters: This helps gauge the scale of the problem and prioritize our response. Expected answer: Around 5-10% of bookings. Impact on approach: A higher percentage would indicate a systemic issue, while a lower one might suggest isolated incidents.

• Considering the nature of car rentals, I'm curious about the types of charges causing discrepancies. Are these primarily related to time overruns, fuel charges, or something else?

Why it matters: Different charge types may point to distinct root causes. Expected answer: A mix, but primarily time overruns and fuel charges. Impact on approach: This would guide our focus on specific areas of the billing system and user behavior.

• Thinking about recent changes, have there been any updates to the pricing algorithm or billing system in the past month?

Why it matters: Recent changes could be directly related to the issue. Expected answer: A minor update to the pricing algorithm was implemented two weeks ago. Impact on approach: This would make us prioritize investigating the recent update and its potential unintended consequences.

• Considering user segments, are these discrepancies more prevalent among new users or experienced customers?

Why it matters: This could indicate whether the issue is related to user understanding or a systemic problem. Expected answer: It's fairly evenly distributed across user segments. Impact on approach: An even distribution would suggest a system-wide issue rather than a user education problem.