PM SYSTEM DESIGN FRAMEWORK: Design Netflix for Series C Startup in APAC

Overview

This framework is calibrated for designing a Netflix-like streaming platform for a Series C startup targeting APAC markets with limited bandwidth infrastructure and regulatory constraints. Focus on PM-level trade-offs optimized for emerging market conditions.

---

PHASE 1: CLARIFYING QUESTIONS (5 minutes)

Product Scope:

• Which APAC markets are primary: India, Indonesia, Vietnam, Philippines, Thailand?
• Content focus: Local original content, licensed international, or both?
• Are we competing with YouTube, local platforms (Viu, iQIYI, Hotstar), or global Netflix?

Technical Constraints:

• Primary device: Mobile (Android-heavy) or smart TV?
• Bandwidth assumptions: 2G/3G prevalent or 4G available?
• Offline viewing: Critical or nice-to-have?

Regulatory Constraints:

• Data localization laws (Indonesia, Vietnam require local storage)?
• Content censorship requirements (China, Vietnam, Thailand)?

---

PHASE 2: USER NEEDS & PERSONAS (5 minutes)

Persona 1: Budget-Conscious Mobile Viewer (60%)

• Demographics: 18-35, urban/semi-urban, limited disposable income
• Need: Affordable entertainment on mobile during commute
• Need: Works on slow internet (3G common)
• Need: Doesn’t burn through data allowance
• Job: “Watch popular shows affordably without buffering or data overage”

Persona 2: Family Subscriber (30%)

• Need: Content for multiple family members (kids, adults)
• Need: Local language content and subtitles
• Job: “Family-friendly entertainment everyone can enjoy together”

Persona 3: Premium Content Enthusiast (10%)

• Need: Latest international originals (Korean dramas, Western series)
• Need: High-quality video when bandwidth allows
• Job: “Premium content worth paying for, even at higher prices”

---

PHASE 3: HIGH-LEVEL ARCHITECTURE (10 minutes)

[Mobile App] <-> [API Gateway] <-> [Content Services]
[Smart TV App]        |                     |
                      |         +-----------+-----------+
                 [CDN Layer]  [Streaming] [Discovery] [User]
                      |         Service     Service    Service
              +-------+-------+    |          |          |
              |               |    |          |      [User DB]
        [Regional CDN]  [Local Edge]  [Transcode]
         (Bangkok,       Cache         Pipeline
          Mumbai,        (ISP-level)      |
          Jakarta)                  [Video Storage]
                                          |
                                   [Multiple Bitrates
                                    + Offline Cache]

Key Components (PM Level):

Regional CDN + Local Edge Cache: Stores popular content close to users

• Why: APAC bandwidth is expensive; latency to US/EU servers kills UX
• PM consideration: Cache hit rate impacts both UX and infrastructure cost
• APAC-specific: Partner with local ISPs for edge caching

Multiple Bitrate Transcoding: Creates 240p, 360p, 480p, 720p, 1080p versions

• Why: Bandwidth varies wildly - 3G users need 240p, WiFi users want 1080p
• APAC-specific: 240p and 360p are primary bitrates (not afterthoughts)

Offline Download: Allows WiFi-only downloads with expiration

• Why: Users download on WiFi, watch on commute to save data
• APAC-specific: Critical feature, not optional

What NOT to Discuss:

• HLS vs DASH streaming protocol details
• Video codec comparisons (H.264 vs H.265 vs AV1)
• Database schema for viewing history

---

PHASE 4: KEY TRADE-OFFS (10 minutes)

Trade-off 1: Freemium + Ads vs Premium Subscription Only

• Options: Free tier with ads | Low-cost subscription ($2-3/month) | Hybrid
• PM Reasoning: APAC price sensitivity is extreme; $10/month Netflix pricing failed in India. Ad-supported models work (Viu, MX Player).
• Recommendation: Hybrid - free tier with ads for top 20% content, premium ($2.99/month) for full library ad-free
• Impact: Larger user base; ad revenue supplements subscriptions; higher complexity

Trade-off 2: Global CDN vs Local Data Centers (Regulatory)

• Options: Global cloud CDN | Local data centers in each regulated market | Hybrid
• PM Reasoning: Indonesia and Vietnam require data localization. Violating blocks market entry.
• Recommendation: Hybrid - AWS local regions (Jakarta, Mumbai) for user data; global CDN for video content
• Impact: Compliant with laws; 20-30% higher cost; slower rollout to new markets

Trade-off 3: High Video Quality vs Data Efficiency

• Options: Standard quality (480p default) | Aggressive compression (240p default) | User-controlled
• PM Reasoning: Data is expensive ($5-10/GB in some markets). YouTube defaults to 240p on mobile in India.
• Recommendation: User-controlled with 240p default on mobile data, auto-upgrade to 480p on WiFi
• Impact: Lower data consumption = higher engagement; lower quality may hurt premium perception

Trade-off 4: Local Originals vs Licensed Content

• Options: Heavy investment in local originals | Focus on licensing | Balanced (80/20)
• PM Reasoning: Local originals drive differentiation but cost $500K-$2M per episode. Licensed content is cheaper but less differentiated.
• Recommendation: Balanced - 80% licensed, 20% budget for 2-3 local originals per year per key market
• Impact: Lower upfront investment; some differentiation; requires local production partnerships

---

PHASE 5: SCALING CONSIDERATIONS (5 minutes)

Geographic Scaling (3 to 15 APAC markets):

• Challenge: Each market has unique regulations, payment methods, languages, content preferences
• Approach: Modular compliance framework; partnership playbook for local payment/telecom
• Trade-off: Slower launches (2-3 months per market) but sustainable

Bandwidth Cost Scaling (TB to PB of monthly data):

• Challenge: CDN costs are dominant expense
• Approach: Volume discounts; ISP peering agreements (free transit for edge cache); aggressive compression
• Trade-off: High upfront negotiation effort but critical for unit economics

---

PHASE 6: SUCCESS METRICS (5 minutes)

Primary Metrics:

• Monthly Active Users (MAU) and growth rate
• Paid conversion rate (free-to-paid for freemium)
• Average watch time per user per month

Technical Health:

• Video start time (target: <3 sec on 3G, <1 sec on 4G)
• Buffering ratio (target: <1% of watch time)
• Download success rate (target: >95%)

Trade-off Monitoring:

• Data consumption per hour by bitrate
• Free-to-paid conversion funnel
• Local original engagement vs licensed content

---

PM vs SWE DEPTH COMPARISON

PM Answer (Just Right):
“Given that many APAC users are on 3G with limited data, I’d default to 240p on mobile data and auto-upgrade to 480p on WiFi. Key trade-off is quality perception vs completion rate - I’d prioritize completion because users will churn if we burn their data allowance. We’d measure data consumption per hour and adjust by market.”

Too SWE:
“I’d implement adaptive bitrate streaming using HLS with a bitrate ladder of 240p at 300kbps, 360p at 600kbps, using H.265 codec. The player would measure throughput every 2 seconds and switch segments based on buffer occupancy using exponential weighted moving average.”

---

REGIONAL CONTEXT: APAC EMERGING MARKETS

Market Characteristics:

• Price Sensitivity: Users accustomed to free or $1-3/month content
• Mobile-First: 80-90% of internet usage on mobile; Android dominates 3:1
• Data Costs: Expensive relative to income ($5-10/GB)
• Payment Methods: Credit card penetration low; mobile wallets, carrier billing common

Key Failure Modes:

• Premium pricing without differentiated content (Netflix struggle in India)
• Ignoring local content preferences
• High data consumption (users churn to save money)
• Regulatory non-compliance (market access blocked)

---

PRACTICE QUESTIONS

• Easier: Design offline download feature for video streaming app in India
• Same level: Design Spotify for APAC emerging markets
• Harder: Design live sports streaming platform for cricket in India

---

Framework Generated by: PM System Design Framework Generator