An Epiphany about Network Effects
16 years ago, in 1999, we founded Tickle, a company most of you don’t remember. It was one of the first viral, user-generated content sites. It offered self-assessment tests and quizzes.
It got pretty big. 100 million registered users when the Internet only had about 600 million users.
But unfortunately, the retention was near zero. People would come to the Tickle website, take the tests for about an hour on average, get viral with their friends, and leave.
Without retention, every few months we had to redesign the product and re-find product market fit. We A/B tested all day. We swam in data and language changes and site releases. We didn’t sleep much or relax. We took no pleasure in our victories because we could see they wouldn’t last long.
We tried adding and subtracting features, and that helped. The product got a bit better every week. But that frenetic feeling wasn’t going away.
During those sleepless months of 2001, a realization hit us. “Our product really doesn’t add any more value to user #100,000,000 than it did to user #1.”
At that moment, the importance of “network effects” was seared into our brains.
We were really good at getting new users, but onboarding those users wasn’t making the product better. With this realization, we started building products that had network effects.
In 2001, we launched a dating marketplace — Tickle Matchmaking, which eventually became LoveHappens. It registered 29 million people who captured each others’ attention with photos and profiles for many months. The more people we had there, the better off other users were.
In 2002, we launched Tickle Social Network – coincidentally the same day MySpace launched — and it grew to 30 million users. These people created content for each other on a daily basis. The more people, the more content, the more valuable for all the members.
In 2003 we launched Grapevine, an advertising marketplace for advertisers to buy ads on user generated content sites like ours, Hi5, Bebo, etc. The more ad units we had, the more valuable to the advertisers. The more advertisers we had, the more valuable Grapevine was to the publishers.
Each of these businesses had a network effect, and they each felt better to run than our initial business. Unfortunately, perhaps due to lack of skill, or perhaps due to running five different businesses inside the company and thus being spread too thin, each of these products fell short of dominating their markets. While we had been profitable for three years, had $32M in revenue doubling every year, and we had plenty of cash… but weren’t out of the woods.
As it happened, in 2004, two months after Facebook launched, Monster offered to buy Tickle for $110M. We took it.
Inside a network effect business
Once we were part of Monster, we experienced the power of a successful network effect.
Monster is a two-sided labor marketplace. Employers on one side, employees on the other. The company was founded in 1994. They dominated their market and had a market cap of $3.5B at the time.
In every element of the business other than the sales team, this was a poorly run company. I don’t think they had changed the website in two years. Poor product, poor customer service, poor strategic decision making, and from what we could tell, a lack of insight into what was about to happen to them because of LinkedIn and others. We tried to get them to let us build a LinkedIn competitor when LinkedIn had less than 1.5 million users, but to no avail.
What stood out was that none of this mismanagement mattered. They had a network effect in place. Like Craigslist, the only feature that mattered was that everyone was there. The buyers found efficiencies in using them and so did the sellers. Both sides of the marketplace kept coming, and Monster kept making money.
So enduring is their network effect, that Monster still posted $770M revenue in 2014, despite hardly changing the product, and despite years of buyout rumors intended to replace the management.
Lesson learned. Network effects were what mattered, and that’s where we were going to focus.
To pursue our newly acquired thesis around network effects, we did several things.
Our first angel investments were all network effect businesses. First was Flickr, which sold to Yahoo. Second, a family social network called Maya’sMom that eventually sold to Johnson & Johnson. Third, Goodreads that sold to Amazon after seven years.
The first board of directors I joined? SecondLife, the virtual world, where every new user add more entertainment and potential revenue to every other user.
Our first purchase was a set of memorable and spellable URL’s for building networks and marketplaces like riddle.com, jiff.com, blue.com, mapper.com, monday.com, healthgorilla.com, nationalgamingleague.com, etc.
This 2013 TechCrunch article chronicles some of the other companies we’ve worked with such as marketplaces Poshmark, Lyft, and Jiff. Since 2013, we’ve been working with companies likeHoneybook and Meerkat. Full list of companies here.
I should also point out a case study failure. We declared Twitter a big deal early. In fact, I tried hard to convince Bill Gurley at Benchmark to invest. Being the gentleman he is, he texted me two years later when Benchmark announced their investment and just said, “We finally did it.” I texted him back saying, “did what?” He sent me the link to the Techcrunch article announcing the investment.
We didn’t invest in Twitter. Why? Because at the time, 2007, I didn’t have an investors’ mindset. I had a builders’ mindset. I was focused on making product. So it simply never occurred to me to email Ev Williams and ask to invest.
Now I’m adopting the investor mindset. That’s part of the thinking behind launching NFX Guild. I’ll post more about that next.