Moreover, We apply [statistical] bands We apply standard statistical analysis to make sure the data looks about right. We compare companies data one to another to make sure the data looks right. Anything suspicious gets challenged, he said.
Yet more than the statistical filter, Glassdoor has a staff that filters all incoming data by hand, he said. The front line of defense is a human nothing replaces a human sitting down, looking at the data.
This human filter is particularly intensive in the Reviews section, where users post feedback about what its like to work at a company including reviews of top execs. (Microsofts hard-charging Steve Ballmer gets a tepid 55 percent approval rating; Ciscos John Chambers easily tops him with 85 percent.)
Ultimately, we review every single review by hand, Hohman said. So everyone passes through a human being and we check for various criteria to make sure it meets community guidelines.
But given that the site will likely receive thousands and thousands of salary numbers and employer reviews, does it have sufficient staff?
Theres probably about a dozen of us, Hohman explained, noting, Were still figuring that out.
What we know is that TripAdvisor still reads everyone single review by hand [the founder of TripAdvisor is on the Glassdoor board]. They made a commitment to data integrity and were making that same commitment to data integrity.
Employee and Employer
As for how Glassdoor will make its money, those details are still being ironed out. (Its good to see that the dotcom spirit of 1999 hasnt died.) Hohman expects to generate revenue from advertising, however, all roads are open, well see what makes sense. At any rate, the site will always be free, he said.
Its likely that the site hopes for some kind of revenue generating deal with employers. Glassdoor is making an effort to get companies involved, including forming an employers advisory panel. And even through launch, we briefed the biggest companies on the site Microsoft and Yahoo, we showed them before launch to get their feedback, Hohman said.
He added: To be really fair and constructive, you need to hear from both sides of this employer and employee.
Perhaps surprisingly, no company has yet complained about this once confidential data being posted online. Not one person [employer] has said, We really wish you wouldnt do this. I dont know maybe they think that and just havent said it, Hohman laughed.
More Data, Better Decisions
Hohman disputed the notion that making this pay data public will somehow harm employees negotiation.
We want to help people make better career decisions, he said. Theres so little good information about what its like to work for a company, what is fair compensation. And sometimes that results in people not making good decisions.
He pointed out that, while Glassdoor posts pay averages (which might group all workers together in an average herd) it also provides the highs and lows. Furthermore, As we get more data, we expect to be able to allow you to dive in and spin that data by, say, years of experience.
In short, providing more data cant hurt workers. Compensation is always a personally negotiated thing, he said. But the more information we put in your hands, the better decision you can make.
More samples from the Glassdoor data:
Yahoo project manager: $83,333 (range: $62k - $105k)
Yahoo senior Unix administrator: $128,500 (range: $113k - $144k)
Yahoo hourly pay: $22/hour
Google site reliability engineer: $95,433 (range: $73k - $110k)
Google product manager: $107,800 (range: $95k - $160k)
Google hourly pay: $21/hour
Microsoft senior consultant: $118,143 (range: $107k - $130k)
Microsoft user experience design: $92,833 (range: $86k - $100k)
Microsoft hourly pay: $41/hour