Friday, February 26, 2010

Cleantech and Cloud Computing Tie the Knot

Quote of the afternoon: "Entrepreneurship is in full throttle. It’s a great time to start a company when there isn’t a shining rainbow in the distance," Warren Packard said this afternoon in Bishop Auditorium.

As the Managing Director at Draper Fisher Jurvetson (DFJ), a venture capitalist company that has managed over $6 billion, Packard knows a thing or two about investing in start-ups. Or 600 things, the number of investments his company has made.

This was refreshing to much of the audience, who was accustomed to hearing what a cacophonous disaster job searching was. I was hooked, ready to hear how these two former GSB students found success along their own separate paths. Yet as is the case with most Stanford students, their career plans converged once again.

Both graduates of '97, Packard and Seamicro CEO Andrew Feldman began talking business again when Feldman stumbled across a great need in the data center industry: power.

"I had no interest in cleantech," Feldman said. "But I saw that data centers like Facebook and Google were such huge power sucks that I thought, 'I gotta solve this.'"

Feldman explained how these giant media companies would trek miles and miles to rivers just to get cheap power. It just didn't make sense for such successful, large companies to have to do this. So with potential customers like Yahoo, Feldman said he saw an opportunity.

He merged the two hot industries of cleantech and cloud computing. The problem? The idea was extremely capital intensive and needed hundreds of millions of dollars. But because Feldman had already identified potential customers, he had already checked off the first step to convincing a venture capitalist to fund his company, Packard explained.

The next steps?

(1. Identifying potential customers, as I already mentioned above)


2. Build a team: This involves lots of breakfasts, lots of lunches with lots of people, Feldman said. He had to pitch this idea as something revolutionary and new. Just a decade ago, data centers' CPUs were all about handling a small number of large tasks. Now, because search engines offer free services, there is an enormous number of small tasks occurring every second. This causes huge energy consumption that can be greatly reduced.


3. Ask for money: Though it’s the most painful part, it doesn't have to be the most difficult part. Packard explained that you shouldn’t classify your company as a "cleantech" company or "cloud computing," but as a company that will solve a customer need.


4. Raising money: This is where DFJ and Packard come in. The most interesting analogy of the afternoon was used when Feldman explained the relationship between start-ups and VCs. His wife works in the emergency room; she treats patients everyday who will remember their emergency room visits for the rest of their lives. Start-ups are like emergency room patients. When you go in for treatment, it’s something you’ll never forget. For medical professionals, or VCs, it’s just another Tuesday. You’ve got to stand out, and VCs will want to invest in your company.


5. Once the VC has committed, they'll take a Board seat. They will line up prospective investors, establish strategy, and jump on your bandwagon.


6. Execution: FINALLY! DFJ's job is to exit the picture and make sure the company can stand on its own two feet. Right now, not the case, but possibly in the future. Feldman explained that he's looking for engineers right now. And though he must pay them $100K less than bigger companies like Cisco and Juniper, he can provide the best experience possible. "I remove every ounce of BS from their lives so they can focus on their job," he explained. "Anything not engineering related is my problem. They don’t get the garbage from big companies in return for a substantial pay reduction."


Like a marriage, Packard said, the journey for a start-up and VC is a wild one, but if there's a spark, something innovative, and a supportive mutual understanding, it can be an ultimate success.


No comments: