The Founders of Tinder and Twitter Knew That to Build Big Buzz, You Have to Start Small How do you build a community around your company? Andreessen Horowitz investor Andrew Chen has a counterintuitive answer: Don't take it to the masses...yet.

What do a dating app, Uber, a nightclub, and Airbnb all have in common? "If your customers come to this product, and there aren't enough other people around, then they're just going to bounce," says Andrew Chen, a general partner at the high-profile Silicon Valley VC firm Andreessen Horowitz, who sits on the boards of Clubhouse, Substack, and a dozen other buzzy tech startups. These companies require what's called a network effect: They become more valuable as more people use them. (After all, nobody hangs out at an empty club.) That means founders need to attract an active and connected base of users quickly, or they may fail to attract anyone at all.

This isn't easy, but it is possible — and one strategy to make it happen applies to any business aiming to foster a community. It's the subject of Chen's new book, The Cold Start Problem. Here, he explains how to attract the masses … by starting with the few.

Related: How to Build a Core Community Around Your Product

When a company is new, how should it begin to build a community?

In my job as a venture capital investor, I see a lot of startups launch by going to big blogs and getting press. As a result, you get a lot of different users showing up — but none of the users know each other. You might get 1,000 users or 10,000 users. But if they're not in the product at the same time to connect with each other, those 1,000 users aren't that valuable. It turns out that the quality of users, and how interconnected they are, matters a lot.

Think about Tinder. If a new user came in, swiped a few times, and then ran out of people, they'd feel like, "The people I want aren't in this app." That's why it's not enough to just do a big announcement. You need to focus on a densely connected population that you can onboard at the same time. I think that's why so many companies, even though they want to get to millions of users and millions of dollars in revenue, start with really small groups. Facebook, Tinder, and Snapchat all started with individual high schools and colleges. Dropbox, Slack, and Zoom started by onboarding individual teams within larger companies. Marketplaces like eBay start in little niches, like collectible cards. Then it spreads from team to team.

I see — because then people find a community they want to engage with, rather than just a bunch of random users they don't care about.

One of the case studies that I use in the book is Theodore Vail, who was president of AT&T in the early 1900s. He wrote an amazing investor update that was so prescient. He said the telephone is useless on its own. It's not even a toy or a scientific instrument. It's only valuable in whom it lets you call, and its value increases with the number of connections. And the point of a product like the telephone — and like a marketplace company like Airbnb or a product like Zoom — is that, even if you nail all the features, it's all about whom you can use that product with.

Related: Building Community Is Good Business

That seems applicable far outside of tech, too.

Yeah. There's a really interesting story about how the first credit card was launched. They picked a specific city — Fresno, CA — and went to the downtown area and basically told the businesses, "Look, we're about to send tens of thousands of these credit card things to your customers. They're going to show up and start trying to use them. Maybe you should accept credit cards." It required some education, but they got more than 300 merchants in the city to pick up these credit card machines. Then they literally mailed live, activated credit cards to people — with no underwriting! — and people just started using them. Then once they proved it could work in one city, they were able to expand to other cities. Or take CrossFit. You feel like you're part of a network, and you do things together, and it creates an opportunity for you to recruit your friends or family. So if you wanted to launch a CrossFit-like program, you could try to do it all over the world all at the same time — but what does it look like to start in one place and just get it working once? If you can prove that it worked once, then it's a lot easier to hypothesize about adding more cities and regions and languages. It's how you bake the idea of a network into your product or service and take advantage of one network at a time.

How do you know when you're ready to expand into another community?

Clubhouse is a great example of this. When we led its Series A, we invested more than $10 million in a company that only had 500 daily active users. That's a really small number. But even with 500 daily active users, you could do an analysis. You could look at the retention curves — so you could ask, of 100 users who join, what percentage are still active a day later? Seven days later? Thirty days later? For a daily-use product, I tend to think that if you still have more than 15% of your users using it on day thirty, that's a good benchmark. If it's a B2B SaaS product, you want to make sure that, after a year, at least half those users are still paying you.

The other big thing that I look at is how fast the product is growing on its own. Is it 20% a month or 30% a month — you know, big double-digit numbers? The best part about products built on networks is that, if they're working, your users and community will tell their friends and bring in new people. You're going to see that top-line number start to move. If that's happening, even if we're talking about a couple hundred users or customers, that's a very strong signal.

Related: The Most Successful Entrepreneurs Are Able to Both Start Small and Think Big

If an entrepreneur doesn't see that growth, how can they figure out what's not working?

That's when you have to figure out, OK, is this because the product features aren't there? In many cases, they aren't.

But the other part could be that you just don't have enough people. When Tinder started, for example, it nailed the features — it built the swiping feature, it added profiles. But in the earliest days, Tinder's founders were just inviting all of their friends. They were like, "Hey, you should try out this app." And they could not get enough people on the app at the same time. So the founders did this whole crazy thing where they threw a birthday party for University of Southern California students, they posted bouncers outside the door, and they got 500 people to install and use the app at the same time. That's what solved the cold-start problem for them.

You have to precisely diagnose the root causes of your problem. That's often very tricky. Create a series of hypotheses and experiments, and start checking them off to make sure that you're validating things before you decide to completely change your product.

Related: How Growing Businesses Can Prioritize Community Involvement

This reminds me of a new fashion designer I just spoke with. She spent a lot of money on ads but has seen almost no traction and is considering giving up. I asked if anything has worked, and she said only one thing did — an ad she ran in a niche publication. But she didn't know why it worked. I told her she needs to figure that out because it could be that she's casting way too wide when her product is actually perfect for a hyperspecific audience she hasn't identified yet.

I think that's totally right. On one hand, yes, maybe this founder needs to figure out more advertising avenues. But the other way is to ask: Is there something special about that community or audience that the founder should actually build a community around? Start to build a network around the idea. Of course, the fashion can be one form of monetization, but there may be many other forms of monetization that this audience would like, including a blog or podcast.

If you own this network, and you're doing a really good job engaging them, giving them content, talking to them, and so on, then they're going to recruit other customers who look and act like them. That's better than the model of spending money on Facebook and Google, where dollars are just flying out the door but you still feel like you're renting an audience as opposed to owning one.

Related: Is the Network Effect Impacting Your Efficiency as an Entrepreneur?