Bebo

Michael Birch's path to Bebo began much earlier, in 1999, when he quit his boring job as a COBOL programmer at an insurance company. His father was an entrepreneur, and Birch was inspired by the idea of building his own thing. He and his wife—also a programmer—decided to make a leap of faith: they'd give themselves three months to make as much money as their combined £300,000 yearly salary ($300k USD). That three months turned into three years of grinding through failed ideas.

Before Bebo, Birch built Birthday Alarm in the early 2000s. He realized something crucial: he was terrible at marketing and sales, so he should only build products that grew virally—where using the product naturally invited others to join. Birthday Alarm hit that magic formula. After moving to Silicon Valley in 2002 and integrating Hotmail and Yahoo address book imports, it grew to 10,000 new users per day. The viral mechanics were elegant: invite friends to enter their birthdays, get reminded of their birthdays, and repeat. Within the first year of monetizing with greeting cards, Birch went from making $10k/month to $10k/day—nearly $4 million in the first twelve months.

Inspired by Friendster, Birch built an early social network called Ringo in 2003 with his brother and friend Morgan Soudin. It reached 400,000 users in three months, but he sold it too early to Tickle for around $700k in stock (which later became worth ~$2M). He was still thinking small. But while at Tickle, he saw them implementing address book scraping at scale—the same viral mechanics he'd perfected on Birthday Alarm. This was the template.

In January 2005, Birch launched Bebo. He reused 80% of Birthday Alarm's codebase, applied the same address book import mechanics, and added photo sharing (inspired by Flickr's success). It took only a couple of days to launch because the underlying architecture was already proven.

Bebo's growth was staggering. With a viral coefficient of 3.5 (meaning every new user brought in 3.5 others), the platform added 350,000 users on its ninth day. At one point, 100,000 new signups came from Singapore alone—a meaningful percentage of the entire country's population. It hit 1 million users in nine days.

However, there was a critical flaw: the original concept of a "self-updating address book" didn't stick. People joined virally but didn't return because an address book with only 10% of your contacts (the portion who actually filled it in) wasn't useful. Birch realized this was an "academic exercise." The solution: shift to being a full social network with photo sharing and social features. That required six months of serious engineering—the longest he'd ever spent on a product—before engagement became genuinely sticky.

One strategic decision that may have cost Bebo: they didn't insist on real identities like Facebook did. Users entered silly names with hearts and symbols, making search nearly impossible. Bebo also launched a year after Facebook and was always in catch-up mode, competing against a better-funded rival with 10x the engineering resources. They raised $15 million in one round, but Facebook had considerably more.

Bebo was eventually sold to AOL for $850 million in 2008. Years later, Birch bought it back for $1 million and sold it again to Amazon. The exit was large, but Facebook had won the social networking war. Still, Bebo's legacy was proving that viral products could be engineered—that constraints (limited resources, no marketing budget) could actually force better product decisions. Birch's thesis: focus relentlessly on products people want to share, not on selling them.