Note that longevity statistics are also affected by growth.

A high-growth company will, by definition, be a short-tenure company. Ironically, it's at shrinking companies that relative tenures often increase, especially as those who cannot (or feel they cannot) compete elsewhere remain.

For a company increasing headcount by 35%/yr, half of all employees will have been hired within the past two years, even with no attrition.

Statistics under conditions of rapid change are ... counterintuitive. And you'd probably want measures that better indicate length against hire date, or at departure date, or similar factors. In some ways this is similar to the challenge of identifying disease mortality during early phases of emergence as not all cases will have reached conclusion and identification of all infections or instances is likely to be incomplete.

Not that I can think of any phenomena where this might be exhibited in the past few years....

A surprising number - sure, google has turnover like everyone else, but in general I am always surprised at how many people have 5+ years of tenure. I work with a number of people where google is the only place they have ever worked, which is downright odd to me.

They take very good care of their staff.