Network Effects: Measure Them, Nurture Them (3 of 3)

Core Metric Framework: The authors developed a list of 16 distinct metrics to measure network effects, moving away from the misconception that these effects are binary (present or absent) and emphasizing their capacity to grow or decline over time.

Leading vs. Lagging Indicators:

• Lagging indicators typically include metrics like the "doubt amount ratio" (often benchmarked at 40–50% engagement) and pricing power (high gross margins allowing for price increases).
• Leading indicators are argued to be more valuable for understanding internal product nuances, specifically:
  • Retention Cohorts: The expectation that recent user cohorts should exhibit higher retention and slower decay curves compared to earlier cohorts as the network matures.
  • Cohort Tension: Early adopter cohorts may show higher retention due to inherent loyalty, creating a tension where the metric must distinguish between founder-driven loyalty and genuine network-driven retention.

Power User Curve Analysis:

• This metric involves plotting a histogram of monthly active users (MAU) against their days of activity within a month (1 to 30 days).
• A successful network effect manifests as a "smile" shape, where the curve shifts right-leaning over time, indicating that a larger proportion of recent users are active every day (30/30) compared to early, more left-leaning cohorts.

Macro Network Strength Measurement:

• The strength of a network is defined by the cost (capital, time, team effort) required for a competitor to build a functional replica.
• Examples of this reconstruction cost hierarchy:
  • Recreating YouTube requires hundreds of millions of dollars to rebuild the library and user base.
  • Launching a ride-sharing platform in a specific city (e.g., Toronto) requires significantly less capital.
  • Launching a niche restaurant delivery platform in a single city district requires the least capital.

Multi-tenanting Dynamics:

• In early stages, users often engage in multi-tenanting (using multiple competing apps simultaneously) due to inconsistent quality or availability.
• As network effects strengthen, multi-tenanting is expected to decline as users develop loyalty to a single platform.

Ultimate Lagging Indicator:

• Pricing power serves as the final validation of network effects, demonstrated when a platform maintains a higher take rate (e.g., 15%) against competitors with lower rates (e.g., 10%) because the platform's inherent value retains users despite the price difference.