How were the strategies for early growth at LinkedIn different from Facebook? originally appeared on Quora - the knowledge sharing network where compelling questions are answered by people with unique insights.

Answer by Matt Cohler, who worked at LinkedIn from early 2003 until he joined Facebook in early 2005, on Quora:

The early growth patterns at LinkedIn and Facebook were very different, even though the goal was the same as for every network-effect or marketplace startup: go from nothing to something which gets better and better as it gets bigger, by tipping into liquidity. All of the great network-effect and marketplace companies I've had the privilege to be involved with have solved this problem, from LinkedIn and Facebook to Instagram, Uber, Snapchat, ResearchGate, Twitter and Tinder. The means to the end have changed over time as the underlying platforms and distribution channels have changed, but the goal is always the same.

Getting to liquidity in a network or marketplace always starts with getting past a classic chicken-and-egg problem: a marketplace or network is only valuable once there are lots of people there, so how do you get people to participate before there are lots of people there? Having a small initial surface area where you're trying to tip a network or marketplace into liquidity always makes this easier; when there are fewer people you need to entice to participate in order to reach a tipping point, the odds of getting a large enough % of those people to do so go way up. And importantly, you can get to that liquidity more quickly when you're starting with a small surface area. Time is a really important and often-overlooked aspect of network growth -- if a network takes too long to tip then you can end up in the self-fulfilling vicious circle of a long bad party.

At Thefacebook we started with American undergraduate schools. These all have the huge benefit of being small, clearly defined spaces, which made it possible for us to tip each one of them very quickly. In many cases it was literally a matter of hours from the time we opened a new school to the time when most people in that school were using the product actively. The specific things we had to do to make this work over and over again were deciding which schools to prioritize opening when; doing a lot of (unpaid!) offline and grassroots local guerrilla marketing; and collecting and bottling-up demand from schools we weren't ready to launch yet then unleashing all of that pent-up demand at once when we were ready to launch a new school. Initially each school was its own island, and then once we had enough schools we started to link the schools together which lead to network effects both within and across schools. It's worth noting that these early chapters at Facebook involved very little true "viral marketing" or "growth hacking" and the early growth at Thefacebook was more linear than exponential as a function of the limitations on who could sign up. I remember being surprised to see how few typical viral mechanisms were built into the product when I first started exploring it in 2004, even by the standards of that time. But those limitations on who could sign up and use Thefacebook were crucial to the core utility of the product for its early users, drove unprecedented engagement numbers, and set a fantastic foundation for the future.

At LinkedIn we didn't have the benefit of a small, clearly-defined initial surface area to go after at the start given the way the product has always worked, so we had to resort to - and in several cases invent - viral mechanisms for growth. In 2003 when LinkedIn launched (many people don't realize that LinkedIn launched almost a year before Thefacebook!) the two primary distribution channels available to us were web search and email. LinkedIn profiles and content weren't showing up meaningfully in search results so we couldn't make much use of SEO. This was because we were very concerned about user privacy and control - in 2003 LinkedIn was actually the first service of any real size where people were putting their full first and last name on a user-generated profile online, and there were originally no pictures on user profiles by design. So email virality was the primary channel available to us for distribution. Like with all exponential functions there was a long initial tail on this viral growth, and Thefacebook actually reached 1,000,000 users much faster than LinkedIn despite not having virality engineered into the product. But once LinkedIn did tip it went from being a curiosity to a service with profound utility. It required a visionary entrepreneur, patient investors, and a team willing to run lots of scrappy experiments to get there.

SEO and email are still both good tools for distribution and growth today, but as the platform has shifted to mobile other things have become at least as important. There's no unpaid mobile distribution platform which is as low-friction and easy as the SEO, email, the MySpace platform (through which YouTube grew) and the Facebook platform (through which Instagram and Pinterest grew) channels were in the Web 2.0 era. If anything this has raised the bar on the importance of finding a way to tip quickly into small networks or marketplaces then expand from there. Uber and Snapchat are both great examples of this: Uber started by focusing on black cars in San Francisco, making that market tip on both the demand and supply side, then launching additional city-level marketplaces from there; Snapchat started with a product that was immediately useful for people using it in a "network" of as few as 2 (or arguably even 1) person thanks to its initial focus on a new form of private communication and private sharing, and grew from there.

This question originally appeared on Quora - the knowledge sharing network where compelling questions are answered by people with unique insights. You can follow Quora on Twitter, Facebook, and Google+. More questions: