Hint: all the questions are about programming. There are no puzzles or tricks, only straightforward problems that require a deep understanding of computer science to solve. And, questions you've read about on the Internet are typically not used.
Google may or may not have higher employee satisfaction, but I sure wouldn't trust Glassdoor to know. This is based on an anonymous 20 question survey of "at least 115 employees" at each company, so you know that one of the two data sets only numbers 115 responses, implying a best-case response rate of 3.8% (Facebook at an estimated 3K employees). There is no information on how the sample set is gathered. Was it self selecting? What type of employees were consulted? Did they span management, or were they all individual contributors? How much can you really learn in 20 questions?
Furthermore, the "best tech company to work for list" that is linked in the article (http://www.zdnet.com/blog/facebook/facebook-named-best-tech-...) is another Glassdoor list, and a highly suspicious one at that. I compared it to the corresponding CNN/Fortune list (http://money.cnn.com/magazines/fortune/best-companies/2012/f...), and out of the top 10 companies on the Glassdoor list, only 3 were on the top 100 Fortune list: Google, Facebook and REI. I've never even heard of 4 of the other 7, including the top 2, and of those 7, 5 aren't even on the CNN/Forbes top 100. On top of all this, this is supposed to be a tech list and it's filled with companies like General Mills, Southwest Airlines, Starbucks, Coach and FedEx, not to mention the fact that the top 2 are also both business consultancy firms.
I could go on, but I think the trend is pretty clear. Glassdoor appears to be basing these ratings on very small (and probably self selecting) samples sizes, and doesn't appear to have a very good means of gathering data from the right companies. Considering it's a relatively new company, this feels like something they want to make big, but they just don't have the brand and sway to implement a well designed study/survey/what-have-you.
Larry Page is a cool guy and the founder of Google, so it's pretty hard to disapprove. It's his company that he built with his own hands. That outweighs any distaste for the concept of the "social multiplier" or whatever it is people hate him for.
I don't think Google's company-wide 10% raise starting in 2011 hurt either.
Approval ratings aren't as useful, as say compared to something like the US Presidental approval ratings: unhappy employees can easily leave (clearly evidenced by a few recent high-profile HN posts), so only the happy remain.
I'd just like to quickly point out that the pie charts for "What gets talked about on the inside" are somewhat useless because they not only don't add up to 100%, but the sum of percentages for the facebook charts doesn't equal the sums of the corresponding google chart (e.g. cons: sum of facebook percentages: 26%, google: 22%)
I agree it's pretty stupid - my best guess is that it could be % of total respondants, not % of total respondants that selected cons. Said another way - 74% of Facebookers didn't list any cons, 78% of Googlers didn't list any cons.
Consider the possibility that googlers dissatisfied with Google tend to leave Google for the bigger and better while there's a strong incentive for a disgruntled facebooker to hang on until his/her options fully vest.
Do you run your own company? It would be interesting to compare side-by-side the benefits of working for Facebook/Google vs. being the CEO of your own startup. Comparing salary, benefits, and perks would be cool, but I'm also wondering if most CEOs find their work more or less interesting and challenging that engineers at these companies do.
Hint: all the questions are about programming. There are no puzzles or tricks, only straightforward problems that require a deep understanding of computer science to solve. And, questions you've read about on the Internet are typically not used.