Instagram and WhatsApp are anomalies. To try and find general tendencies, we should try and cite the typical story, not a three sigma exception case.
Secondly, headcount is a darling of many HR systems. You might cite exceptions with Google or Stripe but your typical corporate environment will not put much effort into fairly evaluating career advancement (e.g., measuring impacts, outcomes, risk) for anyone.
If you are sales or profit focused, lucky you. But outside of that, most HR systems will use simple metrics like "how many people report to you" or "how much budget do you control" to evaluate your promotions and pay. Some will even have literal formulas (e.g., you need 6 reports to go to Director; you need 6 Director reports to go to Executive Director; etc.)
It is done because it is easy and convenient for HR. Unfortunately, obviously, people then try to game the system and over-hire.
You might have to produce an "impacts report" at EOY but that is often HR-theater. Get down to "why" and you will find it came down to how many direct reports you had. Or how much budget you burned.
If your incentive system rewards hiring over results, expect headcount to grow I urge those in governance positions to ensure their organizations have a real impacts assessment system to evaluate employee progress. There is no better way to kill motivation among workers than tell them that impact has no effect.