The Cold Start Problem
ProductStrategy

The Cold Start Problem

Andrew Chen
Read 12 February 2022

Review

Andrew Chen built the clearest framework I have found for reasoning about networked products. He takes one idea, network effects, and constructs a complete architecture around it: five sequential stages, three underlying forces, and a vocabulary that finally makes competitive dynamics in networked markets legible rather than fuzzy. The book isn't empirical; it's a sustained argument made through story. Chen walks through the early survival decisions of Airbnb, the growth mechanics of Slack, and the competitive dynamics of ride-sharing to show how network effects behave in practice, not just in theory. Strategy, competition, product metrics: I came away from this with sharper instincts on all three than any comparable book has managed to give me. Please go and buy the book.

Key Takeaways

The parts worth keeping

The central argument is that network effects aren't a single force. They are a pattern of five sequential challenges every networked product faces, and navigating each one requires a different playbook.

The Five Stages

  1. Cold Start: Build the first atomic network before you can do anything else
  2. Tipping Point: Find a repeatable acquisition strategy and tip the broader market
  3. Escape Velocity: Amplify the three underlying network effects to sustain growth
  4. Hitting the Ceiling: Manage saturation, quality degradation, and hard-side concentration
  5. The Moat: Defend against new entrants who have the structural advantages of a startup

The Atomic Network

An atomic network is the smallest network where enough people are present that everyone sticks around. It's the central concept of the book, the lens through which every other idea is filtered.

Solving the Cold Start Problem

Escape Velocity: The Three Network Effects

The goal at escape velocity is to maintain fast growth while building a sustainable business model. What gets called a network effect is actually three distinct forces operating in parallel:

ForceWhat it does
Acquisition EffectCustomers are acquired more easily as more people join. Viral growth, lower CAC.
Engagement EffectUtility increases as more people join. Density raises retention and usage frequency.
Economic EffectThe business model improves as the network grows: accelerated monetization, reduced costs, better pricing power.

Acquisition Effect in detail:

Engagement Effect in detail:

Economic Effect in detail:

Hitting the Ceiling

As a product reaches scale, growth teeters between expansion and contraction. The negative forces that emerge:

Market saturationRegulatory actionDegradation of marketing channels
Churn from early adoptersMainstream users diluting community qualityNetwork revolts from a concentrated hard side
Spam, trolls, fraud at scaleOvercrowding and broken discovery

Tactics for fighting the ceiling:

The Moat

For networked products, the moat is the effort, time, and capital it would take a competitor to replicate the product and the network. The network is much harder to copy than the product.

GoliathDavid
GoalFighting market saturation and growth slowdownSolving the cold start problem
StrategyAdd new use cases, introduce new audiences, generate profitStart with a niche, skip profitability pressure, focus on top-line growth
ResourcesMore capital, people, and existing productsFewer resources, but no sacred cows
PlaySlower execution, risk aversion, strategy tax: new products must align to the existing businessSpeed and freedom to fail many times; find one atomic market, get investment, grow from there