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

by Andrew Chen

Networked products are fundamentally different from the typical product experience—they facilitate experiences that users have with each other, whereas traditional products emphasize how users interact with the software itself.

It’s why products like Twitter and Zoom and others often seem so simple, and are critiqued as “features not products” that seem trivial at first. They have one magical core experience. Contrast this to traditional products, which often win the “checkmark contest” bake-offs common in enterprise software purchases, but lose in the contest of actual product engagement from their end users.

In recent research by a16z examining the hottest “bottom-up” enterprise startups, most were started by founders from consumer companies like Airbnb, Uber, Yahoo, and so on. The same skills that can create successful networked products in the consumer sector can be applied toward the enterprise categories.

Without being able to build out a network quickly, growth channels like virality are more muted. If Zoom charged every user with no free tier, it might have generated more revenue in the short run, but would likely have needed to spend more on marketing and sales to compensate.

In a platform reinvention, everyone—incumbent or upstart—needs to start over, and faces the Cold Start Problem.

When the network is fully filled out, active, and people are connected in the right way, then the product experience can really shine. This is the Magic Moment, when a product can deliver its core value—whether that’s connecting people for work, entertainment, dating, games, or otherwise.

you start with the opposite of magic, the moments where the network has broken down, and you start solving the problem from there.

The real cost of a zero is not just at the moment where it’s experienced, but rather, the lingering destructive effects afterward.

When a networked product finally starts to generate Magic Moments, it feels really good. Often, this is called “Product/Market Fit”—here’s

You can always feel when product/market fit isn’t happening. The customers aren’t quite getting value out of the product, word of mouth isn’t spreading, usage isn’t growing that fast, press reviews are kind of “blah,” the sales cycle takes too long, and lots of deals never close. And you can always feel product/market fit when it’s happening. The customers are buying the product just as fast as you can make it—or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account. You’re hiring sales and customer support staff as fast as you can. Reporters are calling because they’ve heard about your hot new thing and they want to talk to you about it. You start getting entrepreneur of the year awards from Harvard Business School.26

the Tipping Point. This is where momentum starts to go in your direction, and strategies should be oriented around tipping over entire markets, rather than launching individual atomic networks one at a time. The

From the earliest days, Reid’s theory about professional networks was that, yes, there was a spiderweb of connections, but there was also a hierarchy.

Invite-only mechanics provide a better “welcome experience” for new users as well.

The denser the LinkedIn network, the more likely it was for new users to have a great initial experience.

However, invite-only launches have been a key feature of many products precisely because for networked products, there are huge advantages. It allows the early network to gel as a community, develop a high density of connections, and grow organically via virality.

For networked products, the curation of the network—who’s on it, why they’re there, and how they interact with each other—is as important as its product design. Starting with a deliberate point of view on who’s best for your network will define its magnetism, culture, and ultimate trajectory.

Start with a great “tool”—a product experience that is useful even for one user as a utility. Then, over time, pivot the users into a series of use cases that tap into a “network”—the part where you collaborate, share, communicate, or otherwise interact with other users.

The idea is to initially attract users with a single-player tool and then, over time, get them to participate in a network. The tool helps get to initial critical mass. The network creates the long term value for users, and defensibility for the company.

effect, a tool can be used to “prop up” the value of the network effects curve when the network is small.

The “systems of record” approach comes from building a tool—like GitHub’s source-code version control services—that becomes deeply embedded into a business or workflow, becoming the authoritative and comprehensive version for whatever you’re tracking. For example, GitHub lets individual developers manage their source code, making it a core part of their development environment—this is the tool. The network is then formed by inviting other software engineers to build on the project as well.

it is much easier to spread a tool than a network—after all, the latter suffers from the Cold Start Problem.

The clever part about this is that while it was standard by this point to advertise products in newspapers, by focusing the effort on the hard side of the network—grocers—it bootstrapped the entire network.

The common wisdom was, “If you have a chicken and egg problem—buy the chicken.”

One critique of subsidizing networks in this way is that it’s like “selling a dollar for ninety cents.” Yes, it would be ideal to be able to grow a network with positive unit economics from day one, but sometimes it’s just not possible. Or it might be too slow. A new networked product is better off taking on the risk of subsidizing the network up front, and then improving the economics over time.

Study the pattern of Flintstoning across industries and you’ll see the focus tends to be on replicating the hard side of the network with employees, contractors, and other direct efforts.

High-tech upstarts often employ Flintstoning by throwing employees at it—acting as a traditional brokerage—while slowly automating the most repetitive tasks.

The Flintstoning approach focuses on artificially propping up the hard side of the network with highly manual efforts, whereas the “come for the tool” props it up with software.

once the Cold Start Problem is solved, it’s important to let the network grow and stand up on its own—and turn off Flintstoning entirely.

Creativity is important because there are often brief moments of opportunity that can cause a market to quickly tip, if you try the right idea.

Importantly, he also advocates that B2B startups think about doing consulting to start, treating an initial set of customers as if they were consulting clients. By building the functionality they need on an ad hoc basis, and then generalizing, they have a better chance to hit product/market fit—even though this approach won’t scale.

Hustle and creativity help tip over markets, because each atomic network is not the same. The first, second, and third will likely require slightly different tactics.

Dropbox’s users could be divided into High-Value Actives (HVAs) and Low-Value Actives (LVAs), which was useful as a quality indicator. It could be overlaid into the strategies for marketing channels and partnerships to make sure HVAs were being acquired, not LVAs.

When new products see success and start to scale, it’s often called hitting “Escape Velocity.” The mythology is that the product begins to hockey stick, going up and to the right forever. But it’s not so simple—in reality, the journey isn’t over and instead the focus changes. In this phase, the challenge quickly becomes maintaining a fast growth rate and amplify a successful product’s network effects.

“The Engagement Effect” is what happens when a product gets stickier, and more engaging, as more users join.

“The Acquisition Effect,” on the other hand, is the network effect that powers the acquisition of new customers into your product—in other words, viral growth.

“The Economic Effect.” Network effects can help improve business models over time, in the form of improved feed algorithms, increased conversion rates, premium pricing, and more.

The “Acquisition Effect” is the ability for a product to tap into its network to acquire new customers.

The types of projects that amplify the Acquisition Effect are oriented around viral growth: referral features that reward users when they invite others, tapping into contacts to create suggestions for who to add to an app, and improving conversion along the key moments in the invitation experience. All of these help increase metrics like new user sign-ups, the so-called viral factor of a product, and bring down the cost of acquiring a customer (CAC).

The “Engagement Effect” describes how a denser network creates higher stickiness and usage from its users—it is a more specific form of the classic description of network effects that I covered at the beginning of the book, “the more users that join the network, the more useful it gets.”

The “Economic Effect” is the ability for a networked product to accelerate its monetization, reduce its costs, and otherwise improve its business model, as its network grows.

Here is what’s often called the “Growth Accounting Equation,” which shows how these key metrics relate for active users: Gain or loss in active users = New + Reactivated − Churned Then based on the delta of each time period, you can figure out if you’ll gain or lose active users: This month’s actives = Last month’s actives + gain or loss

The central inputs into a networked product’s growth equation will improve on their own, as a function of the network as opposed to the features of the product—creating an accumulating advantage over time. This is the magic of network effects.

Retention is the most critical metric in understanding a product, but most of the time, the data is not pretty.

As a rough benchmark for evaluating startups at Andreessen Horowitz, I often look for a minimum baseline of 60 percent retention after day 1, 30 percent after day 7, and 15 percent at day 30, where the curve eventually levels out.

What starts as infrequent and noncommittal usage often deepens into daily usage. Luckily, nudging users into more frequent usage can be part of the product design. The key is to target relevant users with messaging or incentives, or otherwise to try out new use cases over time. This moves them from low engagement into high engagement.

These engagement loops are often best visualized, step by step, as a series of actions that tie into each other. Improving any step in the loop benefits all downstream actions.

These churned users are sometimes called “dark nodes.” When they are surrounded by deeply engaged colleagues and friends, even if they’ve been inactive for months, they are often flipped back into an active user.

To amplify the Engagement Effect as it relates to reactivation, the key question to ask is, what is the experience of a churned user?

it’s a combination of big virality projects, lots of little optimizations, and strong retention that ultimately drives big viral factor numbers.

One important insight is that the Acquisition Effect can exist independently of the Engagement or Economic effect. In other words, you can acquire a lot of customers but still have a network that ultimately isn’t sticky.