When Ching-Yu Hu says she's the only "nonrocket scientist" among her company's four co-founders, it's tempting to think she's speaking metaphorically. But before starting Skybox, her co-founders were working on the Google X prize for the moon rover. They are literally rocket scientists.
The three--Dan Berkenstock, Julian Mann, and John Fenwick--were also classmates of Hu's in a 2008 Stanford entrepreneurship class. As part of the class, students had to do the groundwork to form a venture-backed company. Hu says she "literally went up to them and said, 'Hi; I'm going to be part of your team...' They were clearly the three smartest people in the room, and I didn't want to spend my last days at Stanford working on some silly project."
Hu had studied statistics at University of California, Berkeley, and was getting her master's in engineering at Stanford. Before meeting her co-founders, she thought she'd go work on Wall Street. Now, she's handled functions including finance and marketing for a company whose mission is to provide timely high-resolution images of any location in the world using its own fleet of custom-designed satellites.
"We're on a path to launch 24 satellites that will let us see any spot on Earth five times a day," she says. If all goes well, that will happen in about five years. The company is also building a data platform to let it analyze the information coming back from those pictures and integrate it with data from other sources, public and private. If you think that sounds like the kind of data set Google would lust after, you're absolutely right: Skybox agreed to be acquired by the search giant in June for $500 million.
Given the amount of activity in Skybox's industry, the data platform could well turn out to be Skybox's not-so-secret sauce. Urthecast, a Vancouver, British Columbia-based startup, recently released images taken by its cameras, which are bolted to the outside of the International Space Station. Planet Labs is tossing up dozens of supercheap, supersmall satellites, but their resolution will be 3 to 5 meters per pixel, compared with Skybox's 90 centimeters per pixel.
Hu's team started fundraising for Skybox in January 2009. Its pitch detailed its vision for businesses if timely, high-resolution imagery were easier to come by (currently, the U.S. government buys up most of the time available on suitable satellites). Port monitoring, for example, could be transformed. Predictions of crop yields could become much more accurate. Insurance companies and aid organizations could get vastly better, faster information about the effects of natural disasters.
The team nabbed $3 million from Khosla Ventures and has since raised $91 million. Skybox has grown to 130 employees. Its customers can order individual images just before a satellite flies over an area of interest, or they can subscribe to a stream of images, data, analytics, or some combination of the three. Hu says the company has customers and revenue but won't divulge details.
The company designs its own satellites and cameras, then partners with others to build and launch them. Most comparable satellites are about the size of a school bus, with large telescopes and redundant systems. Skybox's are about the size of a mini fridge. "Most of the innovation at Skybox has been about getting a high-performance optic into a small package," says Hu. "The cost of deploying a satellite is directly related to its size and mass."
In turn, Skybox hopes it'll be able to keep costs low enough to appeal both to corporate customers and nongovernmental organizations. Those childhood dreams of having your very own spy satellite are ever closer to coming true.