Let's say you believe you have an opportunity that will double the value of the company, with a 30% probability, or cost the company 10% of it's value. This means it has an expected value of +21%. This is pretty good, and exactly the sort of thing shareholders want from their management. So you increase headcount and pursue the opportunity.
In the 70% scenario when that doesn't work, you have to downsize. Failure is not just possible, it's probable. That doesn't mean that the CEO mismanaged...they may have, I don't know the Dropbox details. But in the scenario where they haven't mismanaged, what do you want? Do you want companies to never take these risks in the first place?