The Cold Start Problem: How to Start and Scale Network Effects

by Andrew Chen

In its classic usage, a network effect describes what happens when products get more valuable as more people use them.

there is a fundamental duality at play—first a physical product, the telephone—and then a second, the network of people and physical wiring that serve to interconnect the phones.

Although the networks don’t own their underlying resources, it’s the connection that matters. The entire ecosystem stays on because the value is in bringing everyone together.

Most products these days are low technical risk—meaning they won’t fail because the teams can’t execute on the engineering side to build the products—but they are generally also low defensibility. When something works, others can follow—and fast.

The systemic value of compatibly communicating devices grows as a square of their number3

Metcalfe’s Law is a simple, academic model that fails the test of real-life messiness.

there was a tipping point—called an “Allee threshold”—where the animals would be safer and thus ultimately grow faster as a population.

The network effect version of this in the technology industry happens when there is “overcrowding” from too many users. For communication apps, you might start to get too many messages. For social products, there might be too much content in feeds, or for marketplaces, too many listings so that finding the right thing becomes a chore. If you don’t apply spam detection, algorithmic feeds, and other ideas, quickly the network becomes unusable. But add the right features to aid discovery, combat spam, and increase relevance within the UI, and you can increase the carrying capacity for users.

Just as crossing the “Allee threshold” is important for a school of sardines to switch from being low/negative growth into a self-sustaining population, when you harvest the sardines more aggressively, you can push them under the threshold.

The Allee effect → The Network Effect Allee Threshold → Tipping Point Carrying capacity → Saturation

studies, I describe an approach that focuses on building an “atomic network”—that is, the smallest possible network that is stable and can grow on its own.

the Acquisition Effect, which lets products tap into the network to drive low-cost, highly efficient user acquisition via viral growth;

the Engagement Effect, which increases interaction between users as networks fill in;

the Economic Effect, which improves monetization levels and conversion rate...

Law of Shitty Clickthroughs, which drives down the performance of acquisition and engagement loops over time, as users tune out of stale marketing channels.

as my friend Naval Ravikant, a noted investor and entrepreneur, has observed: Humans are the networked species. Networks allow us to cooperate when we would otherwise go it alone. And networks allocate the fruits of our cooperation. Money is a network. Religion is a network. A corporation is a network. Roads are a network. Electricity is a network.6

Networks must be organized according to rules. They require Rulers to enforce these rules. Against cheaters. And the Rulers of these networks become the most powerful people in society.

“Slack” was picked to mean “Searchable Log of All Conversation and Knowledge.”

Each workspace might be a business unit or subsidiary, and each smaller team might independently have a ringleader or early adopter who set up the product, invited coworkers in, and started initiating conversations with others.

Generally one side of the network will be easier to attract—this is the easy side of the network. However, the most important part of any early network is attracting and retaining “The Hard Side” of a network—the small percentage of people that typically end up doing most of the work within the community.

This mythology conveniently skips the part of the story when the network is sub-scale and lacks activity. The reality is that new products are often greeted by a nice initial spike of users, but this falls to a trickle as the novelty wears off. Maybe there’s another push, which again goes nowhere. People won’t use their product unless their friends are on it.

Slack works with just 2 people, but it takes 3 to make it really work. There are long-running 3 person groups that are stable—that’s the minimum required to be called a customer.

It’s not enough to sign up, but they also need to be chatting away over time. Eventually once they reach a threshold—for Slack it was approximately 2,000 messages—where they’ll stick around and keep using the product:

Do enough of these analyses, and some interesting patterns will show up—you’ll find the kink in the curve that tells you how much network density is needed to really spike up usage.

For new products, it’s important to have a hypothesis for the size of your network even before you begin.

Ten people using Slack all from the same team is better than ten random people in a larger company. Density and interconnectedness is key.

The solution to the Cold Start Problem starts by understanding how to add a small group of the right people, at the same time, using the product in the right way.

the “atomic network” is the smallest network needed that can stand on its own. It needs to have enough density and stability to break through early anti-network effects, and ultimately grow on its own.

The networked product should be launched in its simplest possible form—not fully featured—so that it has a dead simple value proposition.

The target should be on building a tiny, atomic network—the smallest that could possibly make sense—and focus on building density, ignoring the objection of “market size.”

the attitude in executing the launch should be “do whatever it takes”—even if it’s unscalable or unprofitable—to get momentum, ...

Once Slack could build a dense network that sustained a single team, it could eventually take over an enterprise.

Disruptive technologies are dismissed as toys because when they are first launched they “undershoot” user needs.

The next big thing will start out looking like it’s for a niche network.

Underestimating new products in this way is the number one way to make dumb predictions in the tech industry.

It’s not that product changes are needed—it’s that the network needs to fill out to the point where the people and content are relevant.

Your product’s first atomic network is probably smaller and more specific than you think.

The more users you need to get to an atomic network, the harder it is to create.

Contrast that to when you peanut-butter your efforts across a whole industry or geography—the active parts of the network rapidly dissipate as anti-network effects kick in, because a network of 1,000 random users of Slack will have less retention than 1,000 users all inside the same company.

there is a minority of users that create disproportionate value and as a result, have disproportionate power. This the “hard side” of your network.

they generate what academics often call “cross-side network effects”—when more users in one side of the network benefits the other side of the network.

active contributors represent only 0.02% of the total viewer pool.

Users on the hard side have complex workflows, expect status benefits as well as financial outcomes, and will try competitive products to compare. As a result, their expectations are higher, and it’s difficult to engage and retain them.

across many other platforms, particularly “broadcast” apps where you share videos or photos widely, the value proposition is displaying your status.

The other side of the network will follow. The question is, how? The answer is by building a product that solves an important need for the hard side.

For Uber, in any given market, so-called power drivers constitute 20 percent of the supply but create 60 percent of the trips.

While supply might be easy to get onto the network early on through subsidies, eventually it will become the bottleneck. The hard side of a network is, by definition, hard to scale.

how do you find a problem where the hard side of a network is engaged, but their needs are unaddressed? The answer is to look at hobbies and side hustles.

Sometimes the army is built on people with excess time, but sometimes it is built on people with underutilized assets as well.

The idea is to start with these underserved segments—whose users may not be very attractive customers on their own—and to apply Clayton Christensen’s disruption theory.