Google v Facebook salary inflation riles Big Data startup
Cloudera. It's not the Evil Empire
As Google and Facebook fight over Silicon Valley's top talent, it's a good time to be an engineer. According to one report, when Facebook tried to lure one engineer away from Google, Mountain View counter-offered with a 15 per cent pay raise and $500,000 in restricted stock. And he left for Facebook anyway.
Then, according to other juicy gossip, Google offered a second engineer $3.5 million in restricted stock to reject a Facebook job. And he did.
But all this financial one-upmanship can make life difficult for smaller Silicon Valley startups that rely on similar talent, according to Cloudera CTO Amr Awadallah. Cloudera is what you might call a Red Hat for Hadoop, the open source distributed number-crunching platform based on Google's proprietary back-end infrastructure. Cloudera offers service and support around two of its own Hadoop distros, and it needs the sort of Big Data engineers that both Google and Facebook rely on. Facebook's back-end infrastructure is built, in part, on Hadoop.
Speaking with The Register on Thursday at the Big Data–happy Strata Conference in Santa Clara, California, Awadallah said that as the Palo Alto–based Cloudera works to hire in the Valley, it's seeing engineers receive "substantial" raises from Google and Facebook. "Startups like us just can't compete with that," he said. And this was before Google gave all its employees – worldwide – a 10 per cent raise at the beginning of December.
Cloudera is also seeing extremely high salaries being offered by VMware and Netflix. "In many ways Netflix started the salary-inflation trend by offering employees extremely large salaries and no stock, then giving the employees the choice to use part of their large salary to buy stock in the company at a discount (if they wanted stock)," Awadallah said. "Many employees choose not to buy the stock, and given Netflix's stock price grew 10X over last couple of years, they are probably regretting that decision."
In an effort to become more competitive in the hiring market, Cloudera is opening a new office in the heart of San Francisco's financial district (433 California Street), trying to attract engineers who prefer living in the big city. Facebook's headquarters are in Palo Alto and Google's is in Mountain View – both about 30 miles from San Francisco, but much further away when you consider the epic Bay Area traffic – and although Google runs a San Francisco satellite office, Facebook does not. Zuckerberg and company have instead decided to open an office in Menlo Park, California, just north of Palo Alto – only about five miles closer to San Francisco.
Some San Francisco-based engineers, Awadallah said, have already signed on with the company with the understanding that they will eventually work in the San Francisco office. The office is scheduled to open this March.
If Cloudera is having trouble competing with Google and Facebook salaries, you can bet that other Big Data startups are dealing with much the same dilemma. Cloudera raised $5m in funding in 2009, when virtually no one was investing in startups, and its total funding now tops $36 million. The company has roughly 50 paying customers, including such big names as eBay and Groupon.
The company was founded by an all-star team of Silicon Valley veterans. CEO Mike Olson was previously the CEO of Sleepycat Software, makers of the open source embedded-database engine Berkeley DB, and he spent two years as an Oracle vice president after Ellison & Co. purchased Sleepycat. CTO Awadallah was vice president of engineering at Yahoo!. And chief scientist Jeff Hammerbacher led the Data team at Facebook.
Then, in the summer of 2009, the company hired Yahoo! man Doug Cutting, who founded the Hadoop project. The outfit is no slouch, and if it can't compete with the ridiculous salaries being handed out by Google and Facebook, it can certainly compete in other ways. There's the SF office for one. And we would add that Cloudera is not an Evil Empire built on advertising and invasions of privacy. ®
Sponsored: Hyper-scale data management