Your reading that wrong. Google has 54,000 people working for them and the meadean person working there right now has been with the company just over a year. However, compared to say IBM nobody has been with the company 20+ years and they only had 30,000 people in 2012 so the average is vary influenced by new highers and few long term people to balance it out. Worse there data is stale as people don't keep updating how long they have been with the company.

PS: The real number you want when comparing working conditions is turnover and even that get's influenced by rapid growth.

54K... I'm guessing that's including temps and contractors, which might also include the 1-year average duration.

2 years ago, that was 30k. So, if you assume that the growth is linear, that's 20% of the workforce that has been there less than a year.

According to the article you link, the average tenure is one year for users of a site called PayScale. I've never even heard of it, so they have no idea how long I've worked at Google.

I think you read it wrong.

Its ..

>>The median employee tenure at Google is just more than one year, according to the payroll consultancy PayScale.

He's pointing out that the data comes from PayScale. Most Googlers have never even heard of PayScale, and so the information that is being reported in this article comes from a tiny slice of Google. It's also a biased sample, as someone who goes on PayScale is likely someone who is going to hunt around for the best salary available.

It's dangerous to assume everything you read is correct. You can usually get a hint at how it's biased by reading the article closely.

A number that is skewed by the fast rate of hiring by top-tier companies. When companies are hiring at the rate of Google, Facebook, Apple, etc then even an infinitesimal departure rate of old-timers would set the median at that level.