Median employee tenure is almost entirely dominated by things other than employees leaving. A company that is not hiring will not have any new employees and as a result have a really high median tenure.
The question you really want to ask is: what's the percentage of people leaving each year?
That's also heavily impacted by hiring patterns, though: a company that just hired a lot of people is going to have fewer people leaving (proportionately) after that, since people probably get more likely to leave the longer they've been at the company (though perhaps this is a wrong assumption -- would be interesting to see some data).
You can look at cohorts: of all the people who joined Google in 2013 (eg), what percentage left within one year, what percentage within two years, etc.
The question you really want to ask is: what's the percentage of people leaving each year?