BOOK SUMMARY

The Cold Start Problem - Summary and review

Updated: April 7th, 2024
Published: January 12th, 2024

🚀 Summary: The Cold Start Problem in 30 seconds

How to ignite a network from zero, and use network effects in a product to scale rapidly.

What is The Cold Start Problem about

Three main takeaways

1. The Cold Start Problem - What is network effects, an Atomic Network, the hard side, and avoiding zeros?

A network effect is what happen when a product gets MORE valuable as more people are using it.

☎️ Think of the first telephone. Imagine the increase of the product’s value, when the second person with a telephone joined the network.

An Atomic Network is the smallest network that can stand on its own. It needs to have enough 1. connection density and 2. density to grow on its own.An atomic network is different for different products - Zoom just needs only 2 people in a video call to experience value, Slack decided they need at least 3 users to form a stable network.

The hard side refers to the most difficult side to get on the network. There is a proportion of users that create a disproportionate value. This is the hard side of your network. For Uber it was the power drivers - 20% of the drivers that produced 60% of the trips. For a dating network like Tinder, the hard side is attractive people - that will have a disproportional effect on the value of the network.

Avoiding zeros is about eliminating those “opposite of magic” moments where the users get a terrible experience and often churn as a result. It can be a Uber rider looking for a ride to book, but not having any drivers nearby. Having an active and powerful network reduces the risk of zeros to occur.

2. The Cold Start Theory - A framework for network effects

There are 5 stages of growth for any given network.

Stage 1: The cold start problem 🥶

Most networks fail. The cold start problem is about getting the right people and content on the platform, at the same time. Also, commonly known as the chicken-and-egg problem.Slack isn’t useful until your colleagues are also on the platform. Uber isn’t useful until there are enough drivers, but you also need riders for drivers to want to join.

Stage 2: Tipping point ⚖️

A network is built from many smaller networks. As each small network grows, it becomes easier to scale the next one. As the tipping point is reached, the network has sufficient momentum to build and scale new networks at an exponential rate - think of inflection point where a sliding mass of snow turns into an uncontrolled avalanche. ❄️Meerkat’s Law describes the ecological version of this tipping point, where the value of being a Meerkat 🐵 increases as more animals join the population (as they can warn each other when predators come) - and the “snowball” effect sets into full motion.

Stage 3: Escape velocity 💨

The point where the network is growing at velocity to take off is NOT one single force, rather three forces that need to be up to speed:1. Acquisition effect: lets products tap into the network to drive low-cost, highly efficient user acquisition, via viral growth2. Engagement effect: increases interaction between users as networks fill in3. Economic effect: improves monitization and conversion rates as the network grows

Stage 4: Hitting the ceiling 🚀

In this stage, the network’s growth is stalling and the company is usually in a chaotic state, as it tries to rip itself apart. This stage appears in cycles over and over again as the company breaks through new ceilings of growth.

Stage 5: The moat 🛡️

The final stage where the company uses network effects to defend their grounds. Warren Buffet has popularized the value of looking at companies’ moats when making investments.For networked products like Slack or Airbnb, their software and tech can be replicated quite easily. However, their networks are difficult to clone and thus become major moats.

3. The tipping point - 4 strategies to scale out of the atomic network

Strategy 1: Invite Only

Use FOMO. Provide high touch and high quality onboarding experience for new users. The most connected people tend to be invited earlier - who then will invite other connected people.Clubhouse is a perfect example of this.

Strategy 2: Come for the tool, stay for the network

Attract users with a single player tool. Then, over time, get them to join and participate in the network.

Strategy 3. Paying for launch

Pay for the hard side. This is risky and should only be done at the right time.Uber have done this a lot to acquire and retain drivers.

Strategy 4: Flintstoning

Fred Flintstone used to run himself to power his car. Manually keep the network alive and healthy.This is what Reddit did in the beginning to pick up speed, feeding their communities with the right content and engagement through fake accounts.

Frequently asked questions (FAQs)

Q: Who is Andrew Chen?

Andrew Chen, VC @ Andreessen Horowitz. Led growth team at Uber 2015-2018.