← WhatsApp Interview Insights

WhatsApp·Product Manager·Technical Phone Screen·Senior

SeniorPrefer not to say
May 2026

Summary

Got a product estimation question at WhatsApp, the kind that sounds deceptively simple until you actually try to define what 'median delivery time' even means in a distributed messaging system.

Questions Asked (1)

Q1

Estimate the median time it takes for a message to be delivered on WhatsApp.

Product Analytics & MetricsSystem Design
Author's notes

I started by trying to define 'delivered' which actually helped a lot.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Break down the message delivery process into distinct stages (client to server, server processing, server to recipient) and estimate the time for each stage. Use reasonable assumptions about network latency, server processing, and recipient online status to arrive at a median estimate. Clearly state your assumptions and focus on the median case, not the average or worst case.

Pro tip: Emphasize that the median is dominated by the common case where both sender and recipient are online with good network conditions; acknowledge that offline scenarios create a long tail but don't significantly affect the median. This shows you understand real-world usage patterns and statistical nuances.

1. Clarify the scope and define 'delivery'

Confirm what 'delivery' means: does it mean the message reaches the recipient's device, or is it marked as delivered (double check)? Assume delivery to the recipient's device for this estimate.

2. Map the end-to-end path

Outline the stages: sender's device to WhatsApp server (upload), server processing and routing, and server to recipient's device (download). Consider that messages may be queued if the recipient is offline.

3. Estimate time for each stage

For each stage, estimate typical latencies: e.g., mobile network round-trip time (50-100 ms), server processing (10-50 ms), and push notification delivery (100-500 ms). Sum these for the online case.

4. Account for offline scenarios and median

Recognize that if the recipient is offline, delivery time can be minutes to hours. However, since most users are online when messaging, the median will be close to the online case. Estimate the percentage of messages delivered while recipient is online (e.g., >90%) to justify the median.

5. Synthesize and sanity-check

Sum the stage estimates to get a median time (e.g., 200-500 ms). Sanity-check against known benchmarks (e.g., typical messaging apps deliver in under a second under good conditions).

Key Points to Mention

  • Network latency (RTT) on mobile networks (3G/4G/5G/Wi-Fi) and its variability
  • Server-side processing and message queueing, including WhatsApp's use of Erlang for high concurrency
  • Push notification mechanisms (e.g., APNs, FCM) and their typical delivery times
  • Recipient online/offline status and its impact on delivery time distribution
  • Median vs. average: median is less affected by outliers (offline delivery delays)
  • End-to-end encryption overhead (minimal, but worth noting)

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.