Ticketmaster (Event Booking)

Phase 1 — Clarify & Scope

Interviewer: Design Ticketmaster — an event booking platform for concerts, sports, and live events. What do you want to clarify?

Candidate: Key questions:

1. What's the main complexity here — seat selection and inventory management, or high-throughput flash sales?
2. Do we need a seat map (choose your specific seat) or general admission (section/zone only)?
3. What's the traffic pattern — steady state vs. massive spikes when popular events go on sale?
4. What prevents scalpers from buying all tickets (bot protection, purchase limits)?
5. What's the ticket transfer/resale model?

Interviewer: Both seated events (seat map selection) and general admission. The main challenge is flash sale traffic — Taylor Swift tickets going on sale means 100,000 concurrent users competing for 50,000 seats. Prevent double-booking. Purchase limit: 4 tickets per user per event. Ticket transfer allowed (digital wallet). 10,000 events active at any time.

Candidate: Let me size the flash sale scenario:

$\text{Concurrent users} = 100{,}000 \text{ (during flash sale)}$

$\text{Available seats} = 50{,}000$

$\text{Demand:Supply ratio} = 2:1 \text{ (moderate for popular events)}$

$\text{Ticket selection requests/sec} = 100{,}000 \times \frac{1}{2\text{min wait time}} \approx 833 \text{ req/sec}$

But the real load is the initial surge: all 100K users hitting "buy" simultaneously at ticket on-sale time. This is a thundering herd problem.

Interviewer: What are the core requirements?

Candidate:

• Browse events and view seat maps with real-time availability
• Select seats → hold for 10 minutes (checkout window)
• Complete…