Tokyo Big Sight is a sprawling exhibition hall on the man-made island of Odaiba in Tokyo Bay. In mid-1999, Rigopulos and Egozy were inside at the annual Karaoke Festa trade show, watching their dream die again. They had decided a year earlier that the pot of gold for Harmonix was in Japanese karaoke. So they had leased a costly branch office in Tokyo and put their company's meager cash resources toward developing contacts in the industry. Now, here they were, in the booth of karaoke company Daiichikosho, exhibiting next to a vendor whose product made it possible to order French fries and soda through a karaoke machine. Rigopulos and Egozy thought their gizmo was cooler: a computer joystick glommed onto a real guitar and wired to a karaoke player. Using the joystick, a person could improvise music that sounded like a guitar, while another person sang a popular tune. (The guitar was purely for show, and the joystick jutting out of it when people wore the guitar looked perverse.)
In the booth, they watched as businesspeople politely tried out their invention, nodded, and moved on to test the French fry system. Nobody was having much fun. The smiles seemed forced. Toward the end of the show, Egozy looked around and couldn't find his business partner. Rigopulos was gone. He had realized before Egozy that their idea was doomed and had gone back to the hotel. Karaoke in Japan had nothing to do with improvising a guitar performance, and it wasn't going to. It was another year wasted. About $7 million of investors' money was gone.
It was time to rewrite the business plan again.
Rigopulos calls the early years at Harmonix "the dark ages." For four years, they were making no money, he says. "And to be clear, I don't mean we were earning no money. I mean we had near zero revenue. We were raising money and spending it, building stuff that no one actually wanted to pay for."
They were obsessed with an idea. It started with "Growing Music From Seeds: Parametric Generation and Control of Seed-Based Music for Interactive Composition and Performance," Rigopulos's 80-page master's thesis in 1994 at MIT's Media Lab. The paper had something to do with encoding the essence of music into software.
Rigopulos had grown up outside Boston, messing around with computers and playing drums in a Led Zeppelin/Pink Floyd cover band. As an undergraduate at MIT, he majored in music and co-founded a Balinese percussion orchestra. For his graduate studies, he was assigned to share an office with Egozy, who was a year behind him. Egozy, 36, is a gifted computer scientist, a T-shirt-and-sneakers guy with wire-frame glasses and a grin that seems to say I just changed your grades on the school computer. His family had moved to the U.S. from Israel when he was 12. He is a virtuosic clarinetist who plays in a classical ensemble (though he admits the first album he bought was Van Halen's 1984). The two prodigies clicked.
Some of their peers were working on sophisticated "hyperinstruments" -- hot-rodded cellos and violins and such; one project involved putting computer sensors all over Yo-Yo Ma's cello. "We were interested in the other 99.9999 percent of the population who aren't expert musicians," Rigopulos says.
What if they could take the rush they felt when they played music and package it for popular consumption? They created a computer program that applied basic musical rules to allow anyone to improvise solos by moving a joystick while popular music played. It sounded somewhat funky, and people seemed to have fun with it. "We decided that after we finished our degrees, we had to start a company to try to bring this concept to the world," Rigopulos says. Like so many would-be entrepreneurs with a treasured idea, they were unprepared for a world that didn't really care.
Dude, get up! There's an investor coming!"
It was spring 1996 at the rent-controlled apartment in Cambridge that Egozy shared with two roommates, and Rigopulos had arrived to wake him up. Rigopulos frantically tossed dirty socks and paper into the closet. Egozy folded his futon into a couch and booted up his Mac. A few minutes later, an investor looking to place $25,000 or $50,000 arrived for a demo of The Axe, the name Rigopulos and Egozy had given to their music improvisation software. They can't remember his name; he didn't invest, and more than one Series B financing meeting happened in Egozy's bedroom.
Rigopulos (still living in his parents' attic) was CEO, and Egozy was chief technical officer. They had started Harmonix with $100,000 from friends and family and were now pursuing every other funding contact they could. After losing in MIT's business plan competition, Egozy thought one of the judges' names sounded familiar. He went to the Alpha Delta Phi fraternity house, where he had lived as an undergrad, checked the photos of alumni on the wall, and recognized brother Brad Feld.
"I was completely entranced by Eran and Alex," recalls Feld, now managing director of Foundry Group, a venture capital firm in Boulder, Colorado. "They were incredibly passionate about both music and technology and had tons of ideas." Feld, who had recently sold his own tech company, committed $25,000, made some calls, and delivered a round of $500,000 from angel investors who didn't expect a quick return. And so began the slow beating.
